Upload
others
View
10
Download
0
Embed Size (px)
Citation preview
WHO ear and hearing survey handbook
ISBN 978-92-4-000050-6
© World Health Organization 2020
Some rights reserved. This work is available under the Creative Commons Attribution-
NonCommercial-ShareAlike 3.0 IGO licence (CC BY-NC-SA 3.0 IGO;
https://creativecommons.org/licenses/by-nc-sa/3.0/igo).
Under the terms of this licence, you may copy, redistribute and adapt the work for non-commercial
purposes, provided the work is appropriately cited, as indicated below. In any use of this work, there
should be no suggestion that WHO endorses any specific organization, products or services. The use
of the WHO logo is not permitted. If you adapt the work, then you must license your work under the
same or equivalent Creative Commons licence. If you create a translation of this work, you should
add the following disclaimer along with the suggested citation: “This translation was not created by
the World Health Organization (WHO). WHO is not responsible for the content or accuracy of this
translation. The original English edition shall be the binding and authentic edition”.
Any mediation relating to disputes arising under the licence shall be conducted in accordance with
the mediation rules of the World Intellectual Property Organization.
Suggested citation. WHO ear and hearing survey handbook. Geneva: World Health Organization;
2020. Licence: CC BY-NC-SA 3.0 IGO.
Cataloguing-in-Publication (CIP) data. CIP data are available at http://apps.who.int/iris.
Sales, rights and licensing. To purchase WHO publications, see http://apps.who.int/bookorders. To
submit requests for commercial use and queries on rights and licensing, see
http://www.who.int/about/licensing.
Third-party materials. If you wish to reuse material from this work that is attributed to a third
party, such as tables, figures or images, it is your responsibility to determine whether permission is
needed for that reuse and to obtain permission from the copyright holder. The risk of claims resulting
from infringement of any third-party-owned component in the work rests solely with the user.
General disclaimers. The designations employed and the presentation of the material in this
publication do not imply the expression of any opinion whatsoever on the part of WHO concerning
the legal status of any country, territory, city or area or of its authorities, or concerning the
delimitation of its frontiers or boundaries. Dotted and dashed lines on maps represent approximate
border lines for which there may not yet be full agreement.
The mention of specific companies or of certain manufacturers’ products does not imply that they are
endorsed or recommended by WHO in preference to others of a similar nature that are not
mentioned. Errors and omissions excepted, the names of proprietary products are distinguished by
initial capital letters.
All reasonable precautions have been taken by WHO to verify the information contained in this
publication. However, the published material is being distributed without warranty of any kind,
either expressed or implied. The responsibility for the interpretation and use of the material lies with
the reader. In no event shall WHO be liable for damages arising from its use.
Printed in Switzerland
Contents
Acknowledgements .................................................................................................................................. i
Abbreviations............................................................................................................................................ ii
INTRODUCTION ........................................................................................................................................ 1
DEVELOPMENT AND USE OF THIS SURVEY HANDBOOK ............................................................. 2
Who can use this survey handbook? .............................................................................................. 2
When should the survey handbook be used? .............................................................................. 3
How to use the survey handbook ................................................................................................... 3
Use of survey data ............................................................................................................................... 4
SECTION 1: Planning the survey ........................................................................................................... 5
The planning team .............................................................................................................................. 5
Survey protocol .................................................................................................................................... 7
Survey aims and objectives ............................................................................................................... 8
Estimating sample size ....................................................................................................................... 9
Survey design and methods............................................................................................................ 10
Cluster sampling design ............................................................................................................... 10
Identifying the sample .................................................................................................................. 14
Identifying staff and resource requirements .............................................................................. 16
Developing timelines ........................................................................................................................ 20
Ethical issues ....................................................................................................................................... 21
SECTION 2: Pre-survey preparations ................................................................................................ 23
Training field teams .......................................................................................................................... 23
Pilot and feasibility study ................................................................................................................ 24
Pre-survey visit ................................................................................................................................... 25
SECTION 3: Conducting the survey ................................................................................................... 26
Home visits .......................................................................................................................................... 26
Data collection tools ......................................................................................................................... 28
Care of persons identified with hearing loss or ear problems ............................................... 29
Optimizing participation .................................................................................................................. 30
SECTION 4: Data management .......................................................................................................... 31
Data management plan .................................................................................................................... 31
Minimizing errors in data entry ..................................................................................................... 32
Data analysis ....................................................................................................................................... 32
Data reporting .................................................................................................................................... 34
Hearing loss database ...................................................................................................................... 34
SECTION 5: Quality assurance ............................................................................................................ 35
Prerequisites for a successful survey ............................................................................................ 35
Error sources ....................................................................................................................................... 36
Mitigating errors ................................................................................................................................ 36
SECTION 6: Household roster and data collection form ............................................................. 39
Household roster form ..................................................................................................................... 39
Data collection form ......................................................................................................................... 40
Notes on use of data collection and household roster forms ............................................... 42
SECTION 7: Protocol for rapid assessment of hearing loss ........................................................ 55
Aim of RAHL ........................................................................................................................................ 56
Survey design and sample ............................................................................................................... 56
Examination protocol ....................................................................................................................... 57
Survey team......................................................................................................................................... 58
Equipment required .......................................................................................................................... 58
Missed participants .......................................................................................................................... 58
Data entry and analysis .................................................................................................................... 59
Advantages of RAHL ......................................................................................................................... 60
Limitations of RAHL .......................................................................................................................... 60
ANNEX 1: Sample size estimation and sampling ........................................................................... 61
ANNEX 2: Minimum training requirements for non-qualified audiologists to perform
hearing assessments ............................................................................................................................. 69
Annex 3: Tablet and Smartphone-Based Software Applications for Hearing Assessment 72
ANNEX 4: Exploring funding options – some tips ........................................................................ 74
ANNEX 5: Mobile application and web portal for data collection............................................ 77
ANNEX 6: Protocol for the validation of selected software applications for assessment of
hearing thresholds ................................................................................................................................ 80
i
Acknowledgements
This document is the outcome of a consultative process led by the World Health Organization and
has been reviewed by the following World Health Organization (WHO) staff members: Shelly
Chadha, Alarcos Cieza, Cherian Varughese, Karen Reyes and Tamitza Toroyan.
The conceptual framework and outline of the document were developed in consultation with an
expert group through a series of consultations. Document review and inputs were provided by: Tess
Bright, Sharon Curhan, Suneela Garg, Diego Santana Hernández, Hannah Kuper, Ricardo Martinez,
Islay Mactaggart, Bolajoko Olusanya, Danuk Pallewala, Audra Renyi, Harjit Sandhu, Andrew Smith,
De Wet Swanepoel, Sarah Sydlowski, Peter Thorne and Alejandra Ullauri.
The London School for Hygiene and Tropical Medicine supported this work through hosting of the
consultative meetings and contribution of the section on ‘Rapid Assessment of Hearing Loss’.
Sharon Curhan of Harvard University and Peter Thorne of University of Auckland made a special
contribution through development of a protocol for validation of hearing testing apps. WHO would
also like to acknowledge the developers of the following applications: AudCal; EarTrumpet;
hearTest, Shoebox Audiometry; and uHear, for their support in testing and validation of some of
these apps. The validation studies were undertaken with support of Achal Gulati and Rajiv Dhawan,
New Delhi, India and Peter Thorne and Elizabeth Holt, Auckland, New Zealand.
The final protocol was field tested by Suneela Garg in New Delhi, India and Isaac Macharia in
Nairobi, Kenya. A data management software application, based on the data collection form of this
handbook has been developed by WWHearing to support its implementation.
Financial support for this document has been provided by the International Society of Audiology
and CBM.
ii
Abbreviations
ABR Auditory Brainstem Response
ANSI American National Standards Institute
AOE Acute Otitis Externa
ARHL Age-Related Hearing Loss
ASOM Acute Suppurative Otitis Media
CSOM Chronic Suppurative otitis media
CSOM Chronic Suppurative Otitis Media
dB decibels
dBA A-weighted decibels
dBHL decibels Hearing Level
DHL Disabling Hearing Loss
DHS Demographic and Health Surveys
DPOAE Distortion Product Otoacoustic Emissions
EHCSAT Ear and Hearing Care Situation Analysis Tool
EHES European Health Examination Survey
ENT Ear-Nose-Throat
FB Foreign Body
GBD Global Burden of Diseases
GHDx Global Health Data Exchange
GHLD Global Hearing Loss Database
IHME Institute for Health Metrics and Evaluation
ISO International Standards Organization
kHz Kilohertz
NCDs Non-Communicable Diseases
NGO Non-Governmental Organization
OAE Otoacoustic Emission
OME Otitis Media with Effusion
PIN Personal Identification Number
PPS Probability Proportional to size
PTA Pure Tone Audiometry
iii
RAHL Rapid Assessment of Hearing Loss
RN Random Number
SI Sampling Interval
SOPs Standard Operating Procedures
SPSS Statistical Package for the Social Sciences
SRS Simple Random Sampling
STEPS WHO’s STEPwise approach
SYRS Systematic Random Sampling
TEOAE Transient Evoked Otoacoustic Emissions
TM Tympanic Membrane
UNSD United Nations Statistics Division
WHO World Health Organization
1
INTRODUCTION
Reliable, standardized, population-based data on the prevalence and causes of deafness and
hearing loss are scarce, especially in low- and middle-income countries. A World Health
Organization (WHO) survey undertaken in 20131 revealed that only 30 countries (out of 76
respondents) gathered epidemiological data on hearing loss.
Information on the causes and prevalence of hearing loss essential to raising awareness of
hearing loss and ear disease among WHO Member States, and to developing suitable public
health strategies to address it. Since 1999, WHO’s Ear and Hearing Disorders Survey Protocol
has been used by a number of countries to help conduct epidemiological studies that assess
the prevalence, profile and causes of hearing loss. However, in view of the many advances in
diagnostic technology and software interfaces since 1999, a review of the existing Protocol –
including its methodology, assessment techniques and software for data entry and analysis
was considered necessary.
This survey handbook is the result of that review, and provides guidance for planning and
implementing hearing loss surveys, including information on possible data collection tools.
The survey handbook aims to enable countries – particularly low- and middle-income
countries – to gather data by planning and implementing population-based epidemiological
surveys.
The main uses of data collected by such surveys are:
• to provide an accurate picture of hearing loss prevalence in a given area, which could
be a country or an area within the country (e.g. district or state);
• to provide an overview of the most common probable causes of deafness and hearing
loss in the study area;
• assess global and regional prevalence and trends.
Using this survey handbook for data collection will help to ensure comparability of data
collected through studies conducted in different countries and by different investigators. This
will facilitate the estimation of global prevalence and the examination of hearing loss trends
over time.
1 Multi-country assessment of national capacity to provide hearing care. Geneva: World Health Organization; 2014
2
DEVELOPMENT AND USE OF THIS SURVEY
HANDBOOK
This survey handbook was developed using a consultative process, which was coordinated by
WHO’s team on ‘Sensory functions; Disability; and Rehabilitation’ in Geneva. Following the
decision to update the WHO Ear and Hearing Disorders Survey Protocol, an expert meeting
was held in November 2015 to discuss the scope and nature of revisions and to review newer
available technologies for assessment of hearing.
Based on the outcomes of the meeting, a validation study of software applications designed to
test hearing was undertaken and its results have informed some of the methods outlined in
this survey handbook. An outline of the structure and contents was developed by WHO and
reviewed by experts. The initial draft was prepared by WHO with inputs from consultant
epidemiologists. The expert group reviewed three drafts of the survey handbook.
The final draft was also reviewed by the WHO regional offices. It underwent field evaluation in
India and Kenya in order to assess its usability and feasibility and to identify any practical
difficulties encountered during implementation. On the basis of the inputs and comments
received from reviewers, the survey handbook was modified and reviewed within WHO.
WHO CAN USE THIS SURVEY HANDBOOK?
This document is intended for use by anyone undertaking a study to assess prevalence of
hearing loss in the population. This includes researchers, policy-makers, project managers,
NGOs and others working in the field of ear and hearing care.
3
WHEN SHOULD THE SURVEY HANDBOOK BE USED?
The survey handbook can be used in any situation where the epidemiology of hearing loss is
to be studied in a given population. The aim of such a survey would be to estimate prevalence
of hearing loss and assess its probable causes in the target population.2
Such a survey may be:
• undertaken as part of strategic planning, or to inform public health initiatives;
• initiated by researchers to gather evidence for advocacy on hearing loss, for economic
analysis and for the Global Burden of Disease (GBD) study;
• part of monitoring the implementation of a strategy for ear and hearing care.
Undertaking such a survey at regular intervals (5–10 years) can help to examine trends
in hearing loss prevalence and causes, and to make changes to the strategic plan.
The survey handbook can also be used to guide research studies, such as:
• to determine the prevalence of hearing loss in a defined age group, e.g. adults over
50 years or children aged 5–15 years.
• to assess the prevalence and probable causes of hearing loss in a special
population, e.g. schoolchildren.
HOW TO USE THE SURVEY HANDBOOK
The handbook should be read in full and used to develop project-specific protocols. Each
study needs multidisciplinary expertise from the beginning of the project, and the survey
investigators should include an epidemiologist and biostatistician, as well as an audiologist
and ENT specialist. They should plan the study together and make decisions regarding the
study design and sample; the age range to be tested; the sampling method and sample size;
and data collection, management and reporting. These decisions will in turn determine
logistics such as the personnel required, the type and quantity of equipment needed, and the
time required to undertake and report on the survey and its budget.
2 Refers to the entire group of people to which the researcher intends to generalize the study
findings.
4
However, it is strongly recommended that the epidemiologic methodology, assessment
methods, case definitions and data collection tools be implemented as they have been
outlined in this survey handbook. If these are changed independently, it may not be possible
to use the results for country or regional comparisons.
When required, the study can be supplemented by the collection of additional data that may
be of importance to the target population, so long as this does not interfere with the primary
aim and outcomes of the survey.
USE OF SURVEY DATA
Data on hearing loss prevalence and associated causes, as well as any other information
gathered through this survey, should be disseminated through:
• publication in scientific, peer-reviewed journals;
• preparation and dissemination of a report that can be used for:
o advocacy: the survey results should be graphically presented, and key messages
identified. Advocacy materials, such as infographics/factsheets can be prepared
based on the results. These can be used for raising awareness on this issue among
policy-makers, health professionals and civil society.
o identification of ear and hearing care needs for planning purposes: data collected
will help to determine needs for ear and hearing care in the community and inform
decisions regarding prioritization of services. This information can provide a strong
evidence-base for policy development. The WHO ear and hearing care situation
analysis tool (EHCSAT), the WHO manual on planning for ear and hearing care
strategies and EHC Indicators for monitoring provision of services provide useful
tools for the planning process.3
3 See http://www.who.int/pbd/deafness/news/ear_hearing/en/
http://www.who.int/pbd/deafness/news/ear_hearing/en/
5
SECTION 1: Planning the survey
THE PLANNING TEAM
Once the decision to undertake a survey has been
made, it is important to put together a survey
planning team. The planning team (which may also
be called the steering committee) must include a:
• principal investigator;
• ear and hearing care professionals (ear, nose
and throat (ENT) specialists and audiologist/s);
• public health expert or epidemiologist/s;
• biostatistician/s;
• survey coordinator (responsible for managing
field teams and all field work)
The planning team is responsible for designing the
study; developing the specific survey protocol;
developing the implementation and logistics plan;
developing the budget and seeking funding;
monitoring the study; and quality assurance and
report writing (see Box 1 for key steps in survey
planning). The team should meet at regular intervals
to plan the survey and to monitor its implementation
and evaluate results. All decisions made during the
survey should be documented, together with the
reasons for making them.
The principal investigator should assemble and chair
this team and ensure that it has all the expertise
needed to design, implement, and analyze the survey. The principal investigator should liaise
with government departments and technical consultants and secure funding for the conduct
of the survey. Some possibilities for exploring funding are given in Annex 4.
1. Identify survey planning team members.
2. Appoint survey coordinator.
3. Identify areas (clusters) to be included in the survey
and inform authorities/gain permission, if required.
4. Finalize survey design, methodology and data
collection tools.
5. Develop timeline and budget, and gain financial
approval/support.
6. Finalize study protocol and develop standard
operating procedures.
7. Obtain ethical approval.
8. Establish size and composition of field team and
determine training requirements.
9. Procure required equipment.
10. Train field teams.
11. Conduct pilot and feasibility study.
12. Make pre-survey visit to each cluster.
13. Conduct the study and collect data.
14. Follow up with study group participants, where
required.
15. Data coding, entry and analysis
16. Write survey report.
17. Publish articles in peer-reviewed scientific journals
and share survey results with WHO.
*Many of these steps will occur concurrently rather than
sequentially.
Box 1: Steps for conducting a survey*
6
Figure 1 Structure of the planning team
A full-time survey coordinator is the focal point for day-to-day survey management, in close
liaison with the principal investigator and field teams. As such the survey coordinator must be
part of all planning team meetings, and have knowledge of epidemiology and experience in
field work. Ideally, the survey coordinator should be knowledgeable about hearing loss and
ear diseases. In consultation with the planning team, the survey coordinator will be responsible
for:
✓ preparing the survey protocol and defining standard operating procedures (SOPs) in
consultation with the planning team;
✓ preparing training materials and organization of training;
✓ planning and supervising fieldwork;
✓ arranging and evaluating the pilot test;
✓ preparing advocacy materials to be used by field teams;
✓ supervising data management;
✓ organizing monitoring by the planning team (and external consultants, if required);
✓ reporting progress, problems and any deviations to the planning team.
There may be need for additional support personnel, especially in a large study. If required, an
administrator may be appointed to support the survey coordinator.
PLANNING TEAM
Principal investigator
Ear and hearing care professionals (ENT specialists and audiologist/s
Public health expert or epidemiologist/s
Biostatistician/s
Field team 1 Field team 2 Field team 3
Survey coordinator
7
SURVEY PROTOCOL
The key step in undertaking a survey is to develop a
clear and detailed protocol, which is essential for a
reliable and valid survey. A protocol is the document
that clearly states the survey’s rationale, aim and
objective(s). It describes the survey design and
methodology, and outlines statistical considerations.
Such a protocol helps ensure uniformity in the way
the survey is conducted.
The protocol should be developed by the survey
planning team using a consultative process, and
should include others who need to be involved in the
study, an epidemiologist and a biostatistician. It is also
important to include representatives of the
community to be surveyed as they can help assess
feasibility, highlight specific geographical and
environmental considerations and recommend best
practices for the population under study. Obtaining
political and financial commitment by involving
governments and nongovernmental partners is
important to maximize the use and dissemination of
survey results.
The best way to start is to conduct a protocol
development workshop. All survey protocols must be
approved by an appropriate ethical committee before
implementation.
The parts of a protocol are listed in Box 2. Further
details on a research protocol can be found at
https://www.who.int/ethics/review-committee/format-
research-protocol/en/.
• Project summary
• General information (title, names,
addresses)
• Background and rationale
• References (of literature cited)
• Survey aim and objectives
• Survey design and sample
• Target population
• Survey design
• Sample size
• Sampling strategy
• Methodology
• Safety considerations
• Quality assurance
• Data management
• Expected outcomes
• Dissemination of results and publication
• Duration and timelines
• Problems anticipated
• Project management including survey
team and training
• Ethical considerations
• Other support
• Informed consent forms and
information sheet
• Budget
• Collaborations
• Financing
Box 2: Parts of a survey protocol
8
SURVEY AIMS AND OBJECTIVES
The planning team should first define the aim (or purpose) of the survey, and its objectives.
They must then determine the target population and geographical coverage. Ideally, a survey
should aim to determine the prevalence of hearing loss in all ages in a defined geopolitical
area, such as a country, state, province, district, county, block, governorate, city, town or
village – and assess its probable causes.
In some cases, the survey aim may be to determine the prevalence and probable causes in
hearing loss among a specific subset of the population, such as adults over 50 years or
children aged 5 to 15 years, or schoolchildren.
Once the aim of the study is determined, its objectives can be defined (see Box 3 for sample
survey aim and objectives)
Aim: Determine the prevalence and causes of hearing loss among all people of X
(country/sate/county/district/town/village name)
Objectives:
Primary
o Determine the overall and age-specific prevalence of hearing loss in the population.
o Estimate the overall number of people with hearing loss.
o Estimate the prevalence of different grades of hearing loss (mild, moderate, moderately
severe, severe, profound, and total).
o Determine the age-distribution of hearing loss in the population.
o Determine the prevalence in males compared with females.
o Assess the common causes of hearing loss among different age groups.
o Provide appropriate referrals, treatments or further actions.
Secondary
o To raise awareness about hearing loss in the target population.
o To identify factors that could help prevent hearing loss and inform public health
strategies.
Box 3: Aim and objectives
9
ESTIMATING SAMPLE SIZE
To calculate the sample size for a prevalence survey,
the following steps should be followed, in
consultation with a biostatistician (see Box 4 for an
example of how to calculate a sample size):
Estimate the approximate prevalence of hearing loss in
the target population: this estimate may be based on a
smaller pilot study conducted in part of the
population. It may also be based on available literature
or results of past prevalence studies. Available regional
or country estimates on hearing loss can be used in
absence of previous studies and data. See Annex 1 for
information on sample size estimation.
Determine the required precision of the prevalence
estimate: the precision refers to how far survey
estimates of hearing loss prevalence may lie from the
true prevalence (also known as marginal error). Typical
survey precision may range from 1% to 5%. A value of
1% is preferred for use in these surveys.
Determine the required confidence precision of the
prevalence estimate: 95% confidence intervals are
typical in statistical calculations. The confidence interval represents values for the population
parameter for which the difference between the true parameter and the observed estimate of
the parameter is not statistically significant at the X% level.
Calculate the sample size required for a simple random sample survey (as opposed to a cluster
sample survey) using the required precision, confidence precision, and the estimated
population prevalence of hearing loss.
Prevalence of hearing loss is to be estimated in
a country with:
• a population N=20 million
• estimated hearing loss prevalence
P=5.3%
• required precision d=1%
• required confidence interval 95% (i.e. α
=.05 and z=1.96)
• projected participation rate Pr=85%
Sample size equation: i.e. n=1833 people
Assuming a Pr = .85 i.e. the sample size is
n=1833/.85=2157 people
This assumes that the survey is in ideal
conditions, with no design effect correction
required. However, design effect should be
determined in consultation with a bio-
statistician.
Box 4: Calculating a sample size: an example
10
Estimate the design effect: the cluster sampling design is a complex design associated with a
greater uncertainty than a simple random sample. The design effect is defined as the ratio of
the correct standard error to the standard error computed under the assumption of a simple
random sample. In order to compensate for this, a statistician must estimate the design effect
and the sample size increased in consideration of this factor.
Scientific rigour is essential to ensure reliable results that are representative of the population.
Make a prior prediction of the participation rate: the participation rate refers to the percentage
(proportion) of the sampling frame (eligible population) that will agree to participate in the
survey. The sample size may need to be increased in consideration of the expected
participation rate. It is useful to aim for 90% participation rate or higher in the survey.
However, for the purpose of sample size estimation, it is prudent to be more conservative and
assume an 85% participation rate. Efforts must be made to maximize participation in the
survey (see Section3).
Annex 1 on sample size estimation provides links to useful online resources as well as
available regional estimates that can be used to calculate sample size, as well as
some sample calculations, for reference.
SURVEY DESIGN AND METHODS
Design will undermine the accuracy of its outcome. It is therefore important to plan the survey
in consultation with an epidemiologist and biostatistician. The survey design must aim at
extracting a representative sample from the target population.
Target population: the entire group of people to whom the principal investigator intends to
generalize the study findings.
Sample: all households and participants who are finally included and evaluated as
part of the study.
Sampling unit: all individual items within the sample (e.g. individuals, households). For the
purpose of this survey, sampling unit refers to a household.
Sampling frame: a list of all sampling units that could be selected for the survey.
Source: Health research methodology: a guide for training in research methods. Geneva: World
Health Organization; 2001.
Box 5: Definitions
11
Cluster sampling design
Cluster sampling design is the basic design most likely to be suitable for a population-based
survey carried out using this survey handbook. The objective of this method is to choose a
limited number of smaller geographic areas or population groups (“clusters”) in which simple
or systematic random sampling can be undertaken to select the sample. The term cluster, as
used here, is a collection of sampling units, usually households.
The clusters to be included in the study should be identified using the “probability
proportional to size” method (PPS) (see Box 6 for an example of how to use PPS). The lowest
tier of population groupings or communities are identified prior to cluster selection. These
communities could be villages, urban blocks or sections, or census enumeration areas. Ideally,
detailed maps should be available of these communities, which, at least, show the boundaries
of these communities. Where such maps are not available, they can be generated in
consultation with local leaders.
Communities in which the cluster sampling will be done are identified by PPS. In this method
the likelihood of communities being selected varies according to their size. This ensures that
the sampling procedure is self-weighting and no weighting is required in the analysis. It is
important that the same number of participants/households should be sampled within each
selected cluster.
Identification of sample clusters
Steps to identify clusters:
1. List all communities (e.g. villages or blocks) in the target population. This can usually be
obtained from a census. These communities should be listed with their populations
cumulatively (it is not essential that the population figures are completely up to date,
because the relative size of communities is more important than their absolute size).
12
2. Divide the total number of the target population by the number of clusters required
(see below) to give the sampling interval (SI).
3. Randomly select a number between 1 and the sampling interval (e.g. by using the first
number on a banknote), and the place where this number is located in the cumulative
list of the population identifies the community where the first cluster will be selected.
4. The sampling interval is then added repeatedly to identify the subsequent cluster
locations.
With this method, it is possible that the same community may be selected twice if it has a
population larger than the sampling interval. In this situation two (or more) clusters (subsamples)
should be selected separately within the community.
1. Calculate the sample size.
2. List all the units (e.g. villages or blocks) within the target population and their population sizes.
3. Calculate the cumulative sum of the population sizes and note it alongside.
4. Determine the number of clusters that will be sampled.
5. Divide the total population by the number of clusters to be sampled, to get the sampling interval
(SI): target population/no. of clusters.
6. Choose a random number (RN) between 1 and the SI. The first cluster to be sampled will be the one
which contains the corresponding (cumulative population) number.
7. Subsequent clusters will be identified by repeatedly adding the sampling interval, i.e. RN; RN + SI;
RN + 2SI; and so on.
8. The clusters selected are those for which the cumulative population contains one of the serial
numbers calculated in item 7. Depending on the size of the community (such as village or block), it
is possible that big units will be sampled more than once.
9. Mark the sampled clusters in another column
10. Within each cluster, select secondary sampling units (households) through simple or systematic
random sampling.
See Annex 1 for more information.
Box 6: Steps for cluster sampling using the “probability proportional to size” (PPS) method
13
Number and size of clusters
The number and size of
clusters to be included must
be decided carefully (see Box
7). For a given sample size, there is a
trade-off between increasing the
numbers of clusters, which increases the
precision of the result, and increasing
the size of each cluster,4 which is
logistically easier but reduces the
precision.
• There is no fixed formula to
determine the size of a cluster. Determine the number of individuals to be sampled
from each cluster. In order to ensure that all individuals in the population have the
same probability of selection irrespective of the size of their cluster, the same number
of individuals has to be sampled from each cluster. In practice, the sample size included
from each cluster (cluster size) should generally be the number of subjects that a team
could test in 2 or 3 days. (could be about 150–200 persons). The total number of
clusters would therefore be the sample size divided by the cluster size.
• The number of households to be seen in every cluster will be the cluster size divided by
the average household size. As mentioned above, it is important that the same number
of households should be seen in each cluster. The average household size may be
available from the census or it may be necessary to sample some households to find
out.
• If team size and time allow, it may be possible to increase the number of clusters and
reduce the size of each cluster, and hence increase the precision of the results. Ideally,
the number of clusters can vary between 30 and 50.
4 Size of cluster or cluster size refers to the sample size to be included from each cluster. It does not refer to the total
population of that cluster or community.
To determine the total number of clusters to be included, divide
the total sample size by the cluster size.
To determine the number of households to be seen in every
cluster, divide the cluster size by the average household size (i.e.
the average number of members in each household).
Box 7: Determining the total number of clusters and households
14
• Once the clusters have been identified, a unique number should be allocated to each
cluster. Where there is more than one cluster in a single administrative area, each
should be allocated a unique number. See Annex 1 for more information.
Identifying the sample
The sample is recruited from the available individuals that fall within the sampling frame. It is
important that the size of the sample from each cluster is consistent, i.e. the same number of
people should be sampled in each cluster. After determining the number of households to be
sampled within a cluster, it is important to prepare a map of the cluster, listing each household
and allocating a number to each household (see Box 8).
Next, randomly identify the sample using one of the following approaches:
a) Simple random sampling (SRS): this is an effective method for sampling within clusters.
It can also be used for research studies when the target population is relatively small
and confined to a limited geographical area. It would require that a list of all
households and people residing in the area be made. Each household is then allocated
a number and the sample is then selected randomly from the entire population.
b) Systematic random sampling (SYRS) (see Box 9): this can be applied within a cluster or in
a situation described for SRS (small population and limited geographical coverage). In
this case, after determining the sample size and the sampling interval (N), the first
household is selected randomly, and thereafter every Nth household of the cluster is
1. Note the proposed size of the sample from each cluster (x).
2. Gain information on the average size of households within the clusters (y). This
information is often available through census data.
3. Divide the sample size by household size (x/y) to determine the number of households to
be enlisted in each cluster.
4. Acquire or prepare a map of the cluster, listing each household, and allocate a number to
each household.
Box 8: Steps to identify a sample
15
sampled. While SYRS sampling is likely to be used in most situations, at times it may
mean that certain sections of the population are not suitably represented. For example,
if a district consists of both urban and rural areas, a population-based prevalence
survey must have proportionate representation from both areas. In this case, it would
be appropriate to follow the stratified cluster sampling design.
c) Stratified cluster sampling design: In this design, prior to selecting the clusters, the
population is divided into different strata according to the characteristics under
consideration (e.g. urban or rural, or according to ethnic groups). Each stratum (e.g.
urban or rural) is sampled separately. The results are then weighted by their
populations at the time of analysis.
1. Provide an identification number to each household of the target population is eligible for inclusion in the
study (i.e. the sampling frame).
2. Calculate the number of sampling units (households) to be included by dividing the sample size (for the
cluster) by the average number of persons per household. This gives us the number of sampling units required.
Number of sampling units = Sample size (for the cluster)/average household size
3. Calculate the sampling interval through dividing the total number of sampling units required by the total
number of households within the sampling frame.
Sampling interval (SI) = Number of households within the sampling frame /number of sampling units required
= N
4. Select the first household of the sample population randomly (e.g. the random selection may point to ID 13
as the first household).
5. Select the rest of the sample population by recruiting each Nth subject from the sampling frame (e.g. if
N=100 and 13 is the first house, then select 13th, 113th, 213th subject and so on).
Box 9: Steps for systematic random sampling
16
d) Multi-stage cluster sampling: this method is appropriate when surveying large
populations, e.g. for a national prevalence survey. The entire target population is
divided into large groups, which may correspond to geopolitical divisions such as states
or counties. The required number is randomly selected. Each state/county is then
further divided into smaller groups (such as blocks or villages) and the required number
of these are again identified through the process of random selection. Any number of
such stages can be incorporated until the required cluster level is reached. The sample
can then be identified as described above.
In order to obtain a study sample that is representative of a national or large
subnational population, multi-stage sampling must be incorporated. This type of
sampling is complex, and there is a need to compute corresponding sampling weights
(the number of individuals in the target population that each selected person in the
sample represents). These will help to compensate for unequal probability of selection
and must be done in consultation with an epidemiologist and biostatistician.
IDENTIFYING STAFF AND RESOURCE REQUIREMENTS
Field teams are responsible for survey implementation in the clusters. While the number of
field teams and their composition depend on the survey design and size, each team should
nonetheless include the following:
• A team leader, with the overall clinical and managerial skills necessary to coordinate
and supervise the team’s activities.
• At least one member with specialized knowledge of ear diseases and hearing
assessment (e.g. a trained physician, audiologist or ENT specialist). This person may also
be the team leader.
• Ear and hearing care worker/s: these individuals will undertake the ear examination and
assessment of hearing loss in the field. Such persons must be trained, if required, to
conduct the tests allocated to them.
17
• Field worker/s: these individuals will be responsible for helping with enumeration,
sampling, follow-up and local interaction. It is important that field workers be recruited
from within the community.
• A driver.
The same team leader and ear and hearing care workers should be responsible for carrying
out the survey across multiple clusters. However, field workers should be recruited locally
within each cluster, as that will ensure that there is a local coordination and know-how. It is
best if a trained ear and hearing care worker is teamed up with a field worker who has local
knowledge. It is ideal to have at least one person in the team who can converse in the local
sign language.
Field teams must be fully trained to conduct the survey according to the survey protocol,
including identification of households; induction of participants; ear and hearing examinations;
referrals; and raising awareness on ear and hearing care. These tasks should be duly
considered when planning the budget.
A list of equipment for conducting the survey
and a detailed budget should then be
developed. The budget should consider various
elements, including:
• Staff costs: remuneration and per diems
to be paid to all staff, including local staff
hired from the community.
• Equipment, supplies and consumables
(see Box 10):
o Data collection tools: otoscope;
tympanometer; optoacoustic
emissions OAE machine;
audiometer or alterative hearing
testing software application; other
tools, such as ear probes, forceps
etc.
✓ Otoscopes
✓ Tympanometers
✓ OAE machine
✓ Audiometer or an alternative hearing
testing software application with the
required hardware (device and
headphones)
✓ Other tools, such as ear probe, forceps etc.
✓ Disposables (cotton wool, syringes,
disinfectants)
✓ Ear-drops: antibiotics, wax melting
✓ Antibiotics
✓ Laptop or other devices
Box 10: Equipment for conducting a survey
(one set per field team)
18
o Disposables required for examination, including cotton wool, syringes etc., and
materials to clean/disinfect equipment.
o Medicines and ear-drops for basic treatment of ear conditions, when required.
o Laptop or tablet for data management (if appropriate).
• Printing costs: costs of printing data collection forms, information and consent
documents, survey referral forms, and advocacy materials.
• Training, accommodation and subsistence costs: for training of ear and hearing care
workers and sensitization of field workers.
• Transport costs: for field team, and for planning team members to monitor field
activities from time to time. Transport may also be needed for participants in cases
where hearing tests are not carried out in homes but at a central facility (see Section 3,
subsection “Data collection”).
• Workshops and meetings for the planning team and field teams.
• Data management and analysis.
• Report preparation and dissemination, including reporting back to the community
19
.Budget (proposed categories)
Staff costs
Training and sensitization
Advocacy material preparation
Printing
Workshops and meetings
Data management and analysis
Report preparation
Dissemination of results
Miscellaneous
Cluster 1 Cluster 2 Cluster 3
Field staff
Training
Accommodation
Transport
Equipment
Disposables
Medicines
Miscellaneous
20
DEVELOPING TIMELINES
After finalizing the survey design and methodology, the timeline for conduct of the survey
should be established. This can be documented as a Gantt chart – a chart that plots dates
against tasks (see Figure 2) If required, the timeline may be adapted or modified following the
pilot study or during the course of the survey.
Figure 2 Gantt Chart Example
Timelines-Gantt chartDuration(in
months)
Plan
Duration%Complete Actual Start
PERIODS
1 2 3 4 6 9 13 14 15 16 17
Project Planning 1 2 1 2100%
Data Collection 2 3 1 475%
Data Cleaning 2 4 2 5100%
Data Analysis 3 8 3 6100%
Report Writing 6 2 7 3100%
Submission 8 8 8 9100%
% Complete
(beyond plan)
ACTIVITY PLAN STARTPLAN
DURATION
ACTUAL
START
ACTUAL
DURATION
PERCENT
COMPLETE
Actual (beyond plan)
21
ETHICAL ISSUES
Ethical approval must be gained from a competent authority prior to starting the study. Ethical
matters should be carefully discussed by the steering committee during protocol development
and all issues identified must be documented and addressed. All local rules and regulations
must be followed. Common ethical considerations include the following:
• Discuss the process and procedures with the local community prior to conduct of
survey. Seek the agreement of community leaders and of competent regulatory
authorities (if any) for this purpose.
• Informed consent must include:
o an information sheet provided to all participants. This sheet must clearly provide
all information about the study, its procedures and the proposed use of study
findings. The information sheet must be in the local language and be easy to
understand.
o a consent form in the local language should be made available, with space for
the participant’s signature or thumb impression. In the case of children below
the legal age of consent (as determined by national or local regulations),
consent must be sought from a parent or guardian. The consent form should
also explain and seek permission for any minor procedures, if required (such as
wax or foreign body removal). See WHO sample consent form5 and information
sheet for guidance on how to prepare a consent form (sample form includes
guidance for qualitative studies, studies involving children/minors, and gaining
parental consent). The samples should be adopted and customized to suit each
study’s needs.
For deaf persons using sign language, the information sheet and consent form should be
explained in sign language. If survey participants are unable to read, information should be
verbally explained by a relative or a team member.
• Assent form: for minors (above the age of 7 years), assent should be sought in an
assent form. This should be separate and in addition to the consent form.
5 See http://www.who.int/rpc/research_ethics/informed_consent/en/.
22
• Follow-up and referrals:
o For persons identified with ear disease or hearing loss, suitable treatment should
be initiated, when possible, in consultation with a team member who is
authorized to prescribe such treatment.
o Where a person requires advanced diagnosis or management, he or she should
be referred to a suitable facility, which is accessible to the individual, both
geographically and financially. The referral pathway must be clearly established,
and linkages should be developed with the referral facilities prior to the start of
the survey. Referral slips should be available at the time of assessment. This
should be conveyed to all members of the survey team during training.
o Management of the persons identified with hearing loss or ear diseases must follow local best
practice.
23
SECTION 2: Pre-survey preparations
TRAINING FIELD TEAMS
Field teams should be trained in various aspects of the survey including the following:
• Identification of households within clusters and how to approach household members
regarding their inclusion in the survey.
• Provision of information on the survey and gaining consent from participants.
• Gathering and recording information through household roster forms and data
collection forms.
• Providing information and education on the principles of ear and hearing care.
• Examination of the ear using an otoscope and identification of common ear problems
expected to be encountered during the survey.
• Wax and foreign body removal.
• Assessment of hearing levels.
• Tympanometry and how to record it
• Identification of cases requiring referral, including those in need of urgent medical
attention.
• Referral processes.
While both health workers and field workers must be trained in in steps 1 to 5, only health
workers expected to perform ear and hearing examination should be trained in steps 5 to 10.
When training in ear and hearing examination, it is important to ensure that trainees’ findings
are accurate (i.e. are the same as those that the trainer would produce) prior to authorizing
trainees to conduct examinations as part of the survey.
The WHO Primary ear and hearing care training resource (intermediate level) can be adapted
to suit training needs. Training must include use of the specific ear and hearing assessment
device/s that are to be used under the survey protocol. Any cultural sensitivities in cluster
areas must be considered during training. Field workers must be recruited locally and
sensitized to the needs and objectives of the survey. They should be trained in use of
awareness materials and to prepare them to interact with study participants. See Annex 2 for
more on training field team.
24
PILOT AND FEASIBILITY STUDY
Ideally a feasibility study should be carried out after the sample clusters have been identified,
the protocol developed, equipment procured, and training undertaken. However, a small
study may have to be done earlier in cases where this is required to calculate the sample size.
This can have several purposes: to test the methodology, equipment, logistics and timing of
each part of the survey and to familiarize the staff with conducting the survey. Prevalence
obtained in the pilot may be useful to calculate the sample size if there is no other source for
this information. Issues identified during the study should be documented and can be used to
adapt and finalize the protocol. Feasibility study considerations include the following:
• It can be undertaken at any place within the target population, including a with a small
number of participants, approximately equal to the size of one cluster.
• The feasibility study should be operated with the same number of staff that will
participate in the actual survey field operations.
• All data collection tools and data management protocols should be tested during the
pilot study, but data collected should not be used in the final survey analysis.
• Supervisory staff and trainers should be on hand to provide direct monitoring and in-
the-field training, if required.
• The protocol should be adapted after the feasibility study to accommodate any lessons
learned.
• Individuals who participate in this are not eligible for the full study.
Following the feasibility study, the planning team must finalize the survey protocol before its
implementation.
25
PRE-SURVEY VISIT
Before starting the survey in any location, one or two members of the team should visit the
cluster. They should undertake the following activities during the visit:
• Sensitize the community (through its leaders, whenever possible) and explain the
purpose and procedures of the survey and benefits for the community. This will help to
obtain consent and cooperation of the community for optimum participation.
• Map the location of potential referral facilities for those identified during the survey as
needing additional care.
• Assess the situation regarding suitability of the area, including its accessibility during
different seasons.
• Locate a suitable (quiet and central) site for conduct of ear and hearing assessment
(see Section 3, subsection on “Data collection tools”).
• Obtain a list of households and (if possible) identify households using random
selection.
• Identify field workers for the field team and train them (see Section 2, subsection on
“Training field teams”).
• Secure accommodation for the field team, if required.
• Most surveys are likely to have multiple sites. Each site should be visited before the
start of the survey
26
Figure 3 Steps for conducting the survey
SECTION 3: Conducting the survey
HOME VISITS
The field team should determine which days and the time of the day that is most suitable for
visiting families and conducting the examination. They should prepare a list of households to
be visited, their allocation among the field workers, and an itinerary, to ensure that all
households are visited.
Based on the selection process, the field teams should make a list of all households included in
the survey. Each household should be visited by a small team of two: one ear and hearing care
worker and one field worker.
During this visit, they should inform family members of the purpose and procedures of the
survey and seek their consent for participation. The Household Roster (part of the data
collection form that aims to gather relevant basic information on the households and
participants included in the sample) should be filled in during this visit.
Medication/referral if indicated
Tympanometry
Wax, foreign body removal
Ear examination, otoscopy
5 years and above audiometry0-4 years OAE testing
Data collection form: Part A and B
Registration at the test site
Home visit: Information form and consent form
List household members in advance
27
Families should then be guided to the central location where the ear and hearing examination
will take place and given a date and time for visiting the location.
DATA COLLECTION
It is important that all household members come to the central testing location, and that anyone
who cannot come is either seen later or visited by the team at home (e.g. if the subject is not
mobile). A list of household members, obtained from the census or compiled locally, is essential
to check the names and numbers of those attending and those who fail to attend. If culturally
appropriate and not perceived as unethical, a small incentive in the form of a useful gift item or
monetary compensation may help ensure participation in the survey.
Every attempt possible should be made to find non-attenders. It may be necessary for the team
to return later, e.g. in the evening or at the weekend if that is feasible. In any case, non-attenders
and subjects who refuse to be tested should have a data collection form completed so that they
can be entered into the analysis in order to calculate the non-response rates.
Data collection should be undertaken at an accessible, reasonably quiet location (test site) within the community,
where two or more rooms are available for the survey team. The following steps are recommended:
• The availability and location of such a place should be ascertained during the pre-
survey visit.
• The place selected should be accessible for all, including older adults and people with
disabilities.
• Ideally the place should be centrally located within the community from which the
sample is drawn. However, in case of a large or spread-out area, it is possible that
arrangements for transporting subjects to the test site will have to be made.
Alternatively, it may be required to identify more than one location in the cluster.
• It is important that the test site is comfortable and has place for a registration desk, for
ear examinations and tympanometry, and a separate room where audiological testing
can be undertaken.
As the ambient noise level at the time of audiological testing is required to be below 40 dBA –
a factor that should be checked during the pre-survey visit as well as at the time when tests
are being performed. Validated smartphone-based applications can be used to check the
ambient noise level.
28
DATA COLLECTION TOOLS
Gathering information on, and clinical assessment of, the study group can be undertaken
using the following tools.
The data collection form and household roster form are the basic tools that must be filled in
completely. A standardized data collection form is provided in Section 6 of this survey
handbook and should be used as such. If additional information is required, the relevant
questions or examinations can be added to the data collection form as an additional section.
Instructions regarding its use are provided along with the form in Section 6. Data collection
includes:
• Demographic information and brief history about ear and hearing problems and use of
hearing devices.
• Hearing assessment should be done before ear examination and prior to undertaking
any intervention such as wax or discharge removal.
• 0–5 years of age: OAE testing should be used, combined with tympanometry.
DPOAE/TEOAE test should be used and evaluated using set criteria (see data collection
form, page 40for details)
• Children above 5 years of age: audiological evaluation using audiometry undertaken in
a quiet room or through use of portable audiometry or validated software applications
(see Annex 3).
• In cases where wax, a foreign body or ear discharge is obstructing the ear canal and is
removed, the hearing assessment should be repeated following the removal and also
recorded. NOTE: pre-removal thresholds should be used for estimation of prevalence.
Ear examination: the external ear (auricle and external auditory canal) should be examined
for any abnormalities. The ear canal should be inspected and then examined with the use
of an otoscope. Where feasible, video-otoscopy should be done, and the images of the
tympanic membrane captured. These can be later discussed with the ENT specialists before
forming a firm diagnosis.
Tympanometry: to assess for middle ear abnormalities.
(Note: Care must be taken to ensure disinfection and sterilization of instruments.)
29
Special requests
At times, individuals from the target population who are not part of the sample may request
to be examined. This may be done as a courtesy to the community and provided as an
additional service. Such persons are not a part of the sample and the results should not be
included in data analysis. If someone is identified with hearing loss or ear disease, due care
must be taken to provide the required guidance/intervention or referral.
CARE OF PERSONS IDENTIFIED WITH HEARING LOSS
OR EAR PROBLEMS
The procedure for follow up and care of persons who are identified as having hearing loss or
ear problems during the survey should be decided at the time of planning the survey and
these should be in line with the standards of clinical practice in the area.
All those who require further evaluation or management should be properly guided and
referred to an appropriate and accessible health facility. It is important that the planning team
considers and plans for these situations. Linkages should be developed with health
professionals (physicians, pediatricians, otologists, audiologists) who can help to provide
suitable care for those in need of it. Where financial resources are available, surgical services
and hearing devices should be provided as part of the survey activities. When these are not
available, efforts must still be made to link with other organizations and institutions where
such services can be provided to those needing them, at reduced or no cost.
If any persons requiring urgent attention are identified by the examiners, they should be
properly referred and, if possible, taken to the health facilities (with proper consent) to ensure
timely attention.
30
OPTIMIZING PARTICIPATION
It is important to have the best possible participation rate (at least 80%) in order to have a
valid survey result. Optimizing participation rates includes the following:
✓ A culturally sensitive understanding of the local population.
✓ Raising public awareness about the survey and its importance at national, community
and individual levels.
✓ Engaging the community and local authorities in the survey. As far as possible, field
workers should be hired locally in order to improve access to the population.
✓ Testing the community leader(s) first or before the survey and making this information
public, where possible.
✓ Training health workers/ear and hearing ca\re workers, and field workers in the induction
of participants into the survey.
✓ A clear and participant selection and recruitment process. Proper information on the
survey, including benefits and clear instructions as to how it will be carried out, must be
provided to all participants, in a language which they can easily understand.
✓ Multiple attempts should be made to contact survey participants. The preferred method
of contact may vary with the demographic characteristics and may include a home visit,
phone call, letter or email etc.
✓ Flexibility in scheduling the examination by offering evening and weekend appointments
and rescheduling an appointment where necessary.
✓ Locating the test site as centrally as possible to ensure it is accessible for all, including
older adults and people with disabilities. Where the test site is far away from houses of
participants, transport should be arranged.
✓ Where people are unable to visit the test site due to any reason, home visits and
examination at home should be considered as an option.
✓ Treating survey participants with respect at all times, maintaining confidentiality at all
costs. Photographs must never be taken without prior permission of the participants.
✓ Financial incentives or gifts may be considered but should take cultural norms and
ethics into consideration.
31
SECTION 4: Data management
DATA MANAGEMENT PLAN A data management plan is critical to the success of the survey and will have been outlined as
part of the study protocol. One person must be in overall charge of data management (a data
manager) and should be involved in the discussions regarding the study protocol, data
collection forms, data entry and analysis. The main sources of data are likely to be the
household roster forms and the data collection forms.
Proper data management includes the following:
• The data manager must ensure accuracy and consistency in data collection, recording
and entry.
• All data management procedures should be included in the pilot study, so as to field
test and revise them before survey implementation.
A personal identification number (PIN) should be assigned to each participant. Care
must be taken to ensure that each participant receives a PIN and that only one PIN is
assigned per participant, to ensure smooth data entry. The confidentiality of
participants must be respected, and any clinical and personal data should be
anonymized.
• The planning team must discuss and agree on the software to be used for data entry
and analysis.
The data coding and terminologies noted in the data collection form should be
followed, and the data manager should ensure the uniform application of this coding
during data entry (see Section 7 for the Data collection form).
Data should be entered and checked continuously during data collection (preferably by
two different persons). Validation and consistency checks must be made by the data
manager at every step. This should be done by cross checking at least 5% of the entries
against the data collection forms (where the clinical and demographic details were first
noted). These methods should be clearly noted in the protocol.
Data back-up should be undertaken at regular intervals. All essential documents and
files should be securely stored for a pre-agreed period of time.
32
MINIMIZING ERRORS IN DATA ENTRY
The following steps may help reduce data entry errors:
• Forms and registers should be checked for completeness and consistency before the
fieldwork completion in each cluster. If there are missing data or inconsistencies, field
team members should provide clarification.
• When using paper forms, the original forms and registers should be copied and kept as
a back-up. The original forms and registers are sent to the data manager as soon as the
field activities in one cluster are finished and must be promptly entered into the
appropriate data management software.
• Data validation should be undertaken at the time of data entry. This includes checking
for missing or invalid entries such as incorrect dates, numbers such as age outside
plausible ranges, invalid codes and logical consistency checks.
• 5% of the data entries should be cross checked against the original forms by the data
manager. Up to 1% of errors are considered acceptable. If the error rate is above 1%,
data need to be re-entered
DATA ANALYSIS
Data analysis must be undertaken according to pre-determined indicators. These may include:
• Prevalence in the population.
• Prevalence in specific subsets of the population (based on the objectives of the study):
o According to sex: males; females.
o According to age (at last birthday): e.g. 0–4 years; 5–14 years; 15–44 years; 45–60
years; above 60 years.
o According to geographical distribution: e.g. province A; province B.
o According to urbanization: e.g. urban, rural, urban slums
o Other subsets, as may be required by study objectives.
• Profile of hearing loss in the population, i.e.
o Overall prevalence of hearing loss: to estimate this, the average hearing threshold at
0.5, 1, 2 and 4 kHz should be estimated separately in each ear; hearing threshold
33
above 20 dB in the better ear should be recorded as hearing loss, based on the
grades in Table 1 below.
o Grades of hearing loss: mild, moderate, moderately-severe, severe, profound or
complete (see Annex 7 for full classification).
o Hearing threshold above 35 dB is one ear only should be recorded as unilateral
hearing loss.
o High frequency hearing threshold should also be calculated during data analysis.
This is the average of hearing threshold at 4, 6 and 8 kHz, calculated separately for
each ear.
o Frequency of ear diseases: e.g. impacted wax, otitis externa, otitis media (acute,
chronic suppurative, non-suppurative), dry tympanic membrane perforations and
foreign bodies.
o Other parameters, as required by study objectives (e.g. conductive and
sensorineural type of hearing loss, if this has been studied).
• Occurrence and distribution of ear conditions in the study group.
• Correlation of hearing loss with identified ear conditions.
In the case of complex sampling such as in a multi-stage cluster, it is necessary to use the
sampling weights to compensate for the study design (see Section 1, subsection on
“Identifying the sample”).
Table 1: Estimating hearing loss using grade system
Grades of hearing loss Average hearing level in better ear at 0.5, 1, 2 and 4 kHz
on pure tone audiometry (PTA)
Mild hearing loss 20 to <35 dBHL
Moderate hearing loss 35 to <50 dBHL
Moderate to severe hearing loss 50 to <65 dBHL
Severe hearing loss 65 to <80 dBHL
Profound hearing loss 80 to <95 dBHL
Complete or total hearing
loss/deafness
95 dBHL or greater
Unilateral <20 dBHL in the better ear, 35 dBHL or greater in the
worse ear
34
DATA REPORTING
A detailed report should be developed, including the following sections:
• Background and rationale
• Target population and survey area
• Survey design, sampling methods, sampling weights, and sample size
• Data collection tools, definitions
• Methodology of data collection, including details of technology and its calibration
• Detailed codification of the surveyed data
• Results
• Outcomes and suggested actions
All researchers undertaking a prevalence survey are encouraged to:
Submit findings in a peer reviewed scientific journal
Share the results or if possible, the data with the World Health Organization
HEARING LOSS DATABASE All researchers are encouraged to share the data and results of the survey with WHO. The
World Wide Hearing Foundation’s “WWHearing” software application can be downloaded and
used for data collection, and reporting of some basic demographic information and statistics.
Where agreed after seeking permission from researchers, data entered into this application
can also be shared with others, including WHO. This will be used to build an online database.
Details of the software application and database are annexed (Annex 5
35
SECTION 5: Quality assurance
PREREQUISITES FOR A SUCCESSFUL SURVEY
The following features are essential for conducting a successful study:
• Strong commitment and leadership
• Participation of all stakeholders in the protocol development process
• Effective communication with and among the survey team, stakeholders and
community
• Sufficient financial resources
• Feasibility studies to develop and test protocols, procedures and logistics
• Development of a suitable survey protocol
• Availability of technical capacity and suitable training
• Comprehensive standard operating procedures
• Timely procurement
• Strong data management
• Community participation
• Participation of state (national or subnational) authorities
36
ERROR SOURCES
To collect high-quality survey data, the team must consider possible sources of error and try
to minimize these. Possible error sources include the following:
• Coverage errors. These occur when population elements that are not part of the target
population are sampled, or when those that should be part of the sample are not
sampled. Non-response due to refusal or unavailability of the subject also falls into this
category.
• Observational errors. These are caused by errors in making and recording the
observations and can be attributed to observer, respondent or instrument errors. Inter-
and intra- observer error should be documented and corrected during the survey.
• Processing errors. These are mainly clerical errors that occur during coding, entering or analyzing the
data
Information on ways to minimize errors and enhance quality assurance are listed in Section 4,
subsection on “Minimizing errors in data entry”.
MITIGATING ERRORS
The following factors must be taken into account to ensure a good quality survey that
produces results that are representative of the target population. These are:
• The survey team should include representation from all sectors, including ear and
hearing professionals, biostatistician(s), epidemiologist(s) and the community.
• A clear survey protocol should be developed, including clearly defined standard
operating procedures. Sampling units should be clearly defined at all levels within the
survey.
• All field staff should be carefully trained under actual field conditions, in all aspects of
their duties, and all office staff should be trained in data coding and entry.
37
• Pairing each survey staff member with a field team member selected from the
community can help to maximize participation.
• Survey staff should sign each form and enter their survey staff number when they have
finished completing the form. This encourages good work and makes checking records
much easier.
• Inter and intra- observer variations should be measured and quantified. Thus at the
beginning and at several points during the survey, the performance of team members
should be assessed by comparing different team members testing the same subject,
and comparing a particular team member testing the same subject on different
occasions.
• Supervisors should make regular, announced and unannounced checks on
performance of all survey procedures carried out by survey teams.
• Supervisors should make daily checks of the day’s mapping and enumeration, and of
the completed forms.
• A sub-sample of households should be re-checked by supervisors on a regular basis.
• Every effort should be made to find non-responders. There may be a higher prevalence
of hearing loss among non-responders, and if they are not found the results of the
survey could be biased. If subjects are not at home when the team makes the first visit,
a second visit should be made at an appropriate and if possible arranged time to
include these subjects. This may include weekends because working adults may be
away on weekdays.
Re-evaluate 5% of the sample (5% of the population tested). This should include 5% of those who were
identified as having hearing loss as well as 5% of those who did not have any hearing loss, selected
randomly. The hearing test and ear examination should be repeated by an ENT specialist or audiologist
(different from the primary examiner) to confirm the clinical findings.
Measuring observer variations
38
• Equipment used for hearing assessment should be regularly calibrated, as per the
requirements of the device used. When using an audiometer, this should be calibrated
by a laboratory at the beginning of the study. It should subsequently be checked daily
by a team member using self-calibration, against their known hearing levels.
• Background noise levels should be below 40 dBA during audiological testing, to
maintain optimal test conditions in the field.
• Supervisors should provide field and office staff with regular feedback on their
performance, and provide refresher training if necessary.
• Data entry and processing errors should be minimized as noted in Section 4.
39
SECTION 6: Household roster and data collection
form
HOUSEHOLD ROSTER FORM
(To be filled in during visits to participating households)
Cluster number and household number
Name of head of family
Address
Contact number/s
Family member’s name Age (years)
Gender (M / F)
Educational status
Occupation PIN*
*Participant identification number (PIN) (to be allocated)
APPOINTMENT SLIP TO BE HANDED OVER PARTICIPANTS
For the ear and hearing examination, please come:
To (venue): ……………………………….……………………………….
On (date and time) ……………………………….……………………………….
In case of any questions or concerns, please contact:
………………………………. (name of contact person)
……………………………….
………………………………. (contact number of contact person) ………………………………
40
DATA COLLECTION FORM
DATA COLLECTION FORM A. Participant demographic information (TO BE FILLED IN BEFORE TESTING)
Participant details Participant identification number (PIN) (to be allocated)
The PIN will consist of: Country number/study number/cluster number/household number/participant number
Name:
Date of birth: Sex:
Educational status: None, primary, secondary, post-secondary, graduate, post-graduate
Examination status Date/s of testing
B. History (TO BE FILLED IN BEFORE TESTING)
B1
For adults (aged above 15 years): Without the use of hearing aids or other listening
devices, your hearing is: ☐ 1=excellent; 2=good; 3=a little hearing trouble; 4=moderate hearing trouble; 5=a lot of trouble; 6=deaf; 7=uncertain; 8=refused For children (up to age of 15 years): Which statement best describes your child’s hearing
without a hearing aid or listening device? ☐ 1=excellent; 2=good; 3=a little hearing trouble; 4=moderate hearing trouble; 5=a lot of trouble; 6=deaf; 7=uncertain; 8=refused
B6 Do you (does your child) use a hearing device?
1=hearing aid(s); 2=cochlear implant(s); 3=yes; ☐ 4=no; 5=uncertain Other ……………
B7 Do you (does your child) have difficulty in hearing even if using a hearing device? 1=no difficulty; 2=yes, some difficulty; 3=yes, a lot of
difficulty; 4=uncertain ☐
B8 Are you frequently around loud sound (in your workplace or at home)? (“Loud sound” here refers to background sound such that you have to shout to be understood by someone standing an arm’s length from you) 1=never; 2=once a month or less; 3=2–3 times per
month; 4=once a week;5=almost every day ☐
B2 Have you experienced ringing, roaring, or buzzing in your ears or head that lasts for 5 minutes or longer, over the last 12 months?
1=yes; 2=no; 3=uncertain ☐
B3 Do you have history of discharging/draining ears?
1=yes; 2=no; 3=uncertain ☐
B9 How frequently do you use headphones/earphones? 1=never; 2=once a week or less; 3=2–3 times per week;
4=every day ☐
Supplementary questions, in case response to B1 is 3-6
B10 Did you notice the decrease in your hearing after taking any medications?
1=yes; 2=no; 3=uncertain ☐
B4 How long have you had difficulty in hearing? (To be answered when response to B1 is either 3, 4 or 5) 1: since birth; 2=since childhood; 3=since age 18/59; 4=since old age (60+); uncertain; 5=no
difficulty ☐
Supplementary question, in case response to B2 is 1
B11 11.1 11.2 11.3
Does ringing, roaring, or buzzing in your ears/head interfere with your ability to:
sleep? 1=yes; 2=no; 3=uncertain ☐
concentrate? 1=yes; 2=no; 3=uncertain ☐
work? 1=yes; 2=no; 3=uncertain ☐
perform daily activities? 1=yes; 2=no; 3=uncertain ☐
B5
5.1
5.2
5.3
Do any relatives have difficulty in hearing? (To be answered when response to B1 is either 3, 4 or 5)
Brother/sister: 1=yes; 2=no; 3=uncertain ☐
Parent/s: 1=yes; 2=no; 3=uncertain ☐
Children: 1=Yes; 2=no; 3=uncertain ☐
41
C. Hearing examination
C1 Right ear Left ear
OAE test (up to 5 years) 1. Pass 2 Refer 1. Pass 2 Refer
C2 Hearing threshold evaluation in participants aged above 5 years (insert actual value)
Frequency (Hz) 500 1000 2000 4000 6000 8000
Hearing threshold right ear(dB)
Hearing threshold left ear (dB)
Ambient noise level: Screening examiner’s number: Not fully examined in this section: Screening examiner’s remarks: …………………
D. Ear examination findings Code to be entered: 1=yes; 2=no; 3=uncertain Right Left Right Left
D1 Auricle malformation D3 Tympanic membrane
D2 External ear Perforation
Wax Dull and retracted
Wax removed Red and bulging
Foreign body (FB) Normal
FB removed Not seen
Inflammation of ear canal D4 Middle ear
Inflammation of ear canal treated Otorrhoea
Otorrhoea Cholesteatoma
Otorrhoea removed Normal
Fungus (Otomycosis) Not seen
Fungus treated D5 Others:
D6 Tympanometry: Code: 1=A; 2=B; 3=C; 4=As; 5=AD; 6=uncertain; 7=test could not be performed
Screening examiner’s number: Not fully examined in this section: Screening examiner’s remarks:
E. Probable cause of hearing loss (in case of hearing threshold above 20 dB in one or both ears)
Right ear 1=yes; 2=no; 3=uncertain Left ear 1=yes; 2=no; 3=uncertain
E1 Unknown E2 Wax E3 CSOM E4 OME E5 AOE E6 ASOM E7 FB E8 ARHL E9 Other
Examiner’s comments
F. Action needed Code 1=yes; 2=no F1 No action
F2 Medication advised
F3 Refer for further evaluation
F4 Refer for urgent evaluation
42
NOTES ON USE OF DATA COLLECTION AND
HOUSEHOLD ROSTER FORMS
Data collection form
The data collection form is a standardized tool for gathering and recording information during
the survey. The data collection form:
(1) is the data-collection instrument for a population-based survey of the prevalence and
likely causes of hearing loss;
(2) enables data collection in a standard way so that results can be compared amongst
surveys that use this form;
(3) helps gather information on the otological actions needed in order to estimate the
resources required to deal with these problems in the study population.
It is therefore vital that no changes are made to the form when are used in a survey. However,
those conducting the survey may very well decide to gather additional information at the
same time and this can be added as a supplement to this form.
Household roster form
When visiting the household, a Roster form (page no. 43) should be completed, one for each
household. This should include all persons usually living in that household. The personal
identification number (PIN) should be allocated and noted on the Roster Form for each
individual within a household. This same PIN will be inserted into the data collection form (see
description of PIN below). A copy of the data collection form should be handed to members
of the household and provide information regarding place and time of data collection.
The WHO Ear and Hearing Survey Data collection form comprises six different sections, as
follows:
Demographic information
History
Hearing examination
Basic ear assessment
Cause of ear disease and/or hearing loss
Action needed
43
To complete the form, adequate training of survey team personnel is required in advance.
Section A and B need literate and numerate personnel. These sections can be filled in either
during the household visit or when participants report to the test site. Subsequent sections
should be filled in at the test site. Section C needs personnel with audiometric skills to handle
equipment to assess hearing. Section D, E and F should be completed in all cases by
personnel with experience of work as ENT specialists or health workers with higher level
specialist ENT training.
The form must be completed according to the coding instructions. The form is designed to be
easy to fill in by marking the boxes and giving further information where indicated.
When the category “specify” is encountered, a detailed description or opinion of the examiner
should be stated.
During the survey, the availability of services for ear care, such as provision of medication, wax
removal or counselling would be desirable.
A: DEMOGRAPHIC INFORMATION
A unique PIN number should be allocated to each individual. This PIN number will consist of a
series of numbers, filled in boxes A1 to A5, which form a unique identification number for each
participant.
In this section each box should be filled either by a digit or a letter inside a box, according to
the instructions below. The PIN number should be pre-allocated on the data collection forms
to avoid duplication.
The PIN consists of: Country number/study number/cluster number/household number/participant number
Allocating a PIN number
44
NUMBER ITEM INSTRUCTIONS
A.1 Country
number
The UN 3-figure code must be used if data processing is to take place
outside the country. Refer
https://www.nationsonline.org/oneworld/country_code_list.htm
A.2 Study number A reference number to be given to each study conducted within the country.
This code, as well as the country code, should preferably be stamped on
forms prior to data collection.
A.3 Cluster This is a grouping of households, such as a village or an enumeration area.
A different number should be given to each cluster within each of the
administrative divisions.
A.4 Household Household is the smallest group where people live together, such as a family.
The number inserted should identify each household within the cluster. This
number should be the same as that on the Household roster form
A.5 Person The number inserted should identify each person who is a usual resident
within the household. Usual resident is defined as a person having resided in
the household for a total period of 6 months or more. This number should
be the same as that on the Household Roster Form (see description above).
A.6 Name The participant’s name may be filled in for identification purposes but will
not be coded.
A.7 Date of birth This should be entered in the format dd/mm/year. Example: 5 October, 1994
would be 05/10/1994
A.8 Male/Female Mark M or F as appropriate (1=Female; 2=Male)
A.9 Educational
status
Highest educational level should be noted starting from high school
1=high school student; 2=high school graduate; 3=university graduate;
4=post-graduate studies completed; 5=unknown; 6=not applicable (in case
of a child below high school level); 7=school drop out
A.10 Exam status 1=examined, 2=refused, 3=absent
Code 1 (=examined) should be entered if either complete or partial
examination has been performed.
Code 2 (=refused) should be entered if the participant (or parent or
guardian) refuses to have any type of examination or provide any
information. Code 3 (=absent) should be entered if the participant is a usual
resident who is not present during the entire survey period. Even if the
participant is absent or refused, the census section should still be completed
as far as possible and the form submitted for analysis
A.11 Date Date, month and year of this examination should be entered in the format
dd/mm/year.
45
B: HISTORY
The following questions can be used to gather baseline information regarding participants’
hearing status and history of risk factors.
The core questions (B1–3) must be asked of all participants. The supplementary questions (B4–
11) need to be asked as indicated.
These questions should be administered as such. The code (i.e. number 1–8) should be
entered in the box opposite the question in the data collection form. If required,
supplementary questions can be added.
These questions can be asked either during the household visit or when the participants are at
the test site.
CORE QUESTIONS
B1 For adults (above 15 years of age):
Without the use of hearing aids or other listening devices, your hearing is:
1=excellent; 2=good; 3=a little hearing trouble; 4=moderate hearing trouble; 5=a lot of
trouble; 6=deaf; 7=uncertain; 8=refused
For children (up to 15 years of age):
Which statement best describes your child’s hearing without a hearing aid or listening
device?
1=excellent; 2=good; 3=a little hearing trouble; 4=moderate hearing trouble; 5=a lot of
trouble; 6=deaf; 7=uncertain; 8=refused
B2 Have you experienced ringing, roaring, or buzzing in your ears or head that lasts for 5
minutes or longer, over the last 12 months?
1=yes; 2=no; 3=uncertain
B3 Do you have history of discharging/draining ears?
1=yes; 2=no; 3=uncertain
Supplementary questions, in case response to B1 is 3–6
B4 How long have you had difficulty in hearing? (To be answered when response to B1 is
either 3, 4 or 5)
46
1: since birth; 2=since childhood; 3=since age 18/59; 4=since old age (60+); uncertain;
5=no difficulty
B5
5.1
5.2
5.3
Do any relatives have difficulty in hearing? (To be answered when response to B1 is
either 3, 4 or 5).
Brother/sister: 1=yes; 2=no; 3=uncertain
Parent/s: 1=yes; 2=no; 3=uncertain
Children: 1=yes; 2=no; 3=uncertain
B6 Do you (does your child) use a hearing device?
1=Hearing aid(s); 2=cochlear implant(s); 3=yes, other ……………; 4=no; 5=uncertain
B7 Do you (does your child) have difficulty in hearing even if using a hearing device?
1=no difficulty; 2=yes, some difficulty; 3=yes, a lot of difficulty; 4=uncertain
B8 Are you frequently around loud sound (in your workplace or at home)? (“Loud sound”
here refers to background sound such that you have to shout to be understood by
someone standing an arm’s length away)
1=never; 2=once a month or less; 3=2–3 times per month; 4=once a week; 5=almost
every day
B9 How frequently do you use headphones/earphones?
1=Never; 2=once a week or less; 3=2–3 times per week; 4=every day
B10 Did you notice the decrease in your hearing after taking any medications?
1=yes; 2=no; 3=uncertain
Supplementary questions, in case response to B2 is 1
B11 Does ringing, roaring, or buzzing in your ears or head interfere with your ability to:
11.1 sleep? 1=yes; 2=no; 3=uncertain
11.2 concentrate? 1=yes; 2=no; 3=uncertain
11.3 work? 1=yes; 2=no; 3=uncertain
11.4 perform daily activities? 1=yes; 2=no; 3=uncertain
C: HEARING EXAMINATION
A hearing examination should take place at the test site and be carried out by a trained
person. If occluding wax or a foreign body or pus is present, audiometry should be performed
without removing it, and the results recorded in the boxes in Section B of the form. This is in
order to measure the existing level of hearing (before any intervention). (The examiner may
request further audiology after removal for diagnostic purposes, but there is no need to
record the second set of audiological results.)
47
In cases where wax, a foreign body or ear discharge is obstructing the ear canal and is removed,
the hearing assessment should be repeated following the removal and also recorded. However,
the pre-removal thresholds should be used for estimation of prevalence.
As hearing assessment is to take place prior to removal of ear wax or discharge, it is advisable
to use supra-aural or circum-aural headphones with audio-cups and not to insert earphones for
the hearing assessment.
Ambient noise: The testing must be performed in as quiet a room as possible. Ambient noise is
recommended not to exceed 40 dBA measured by a sound level meter or a validated software
application.6 The level of ambient noise should be recorded in the relevant box in section C of the
data collection form. Even if ambient noise is higher than 40 dBA, hearing testing should
continue, and the results, including the level of ambient noise, should still be recorded in the
boxes. All equipment used in testing should be assigned a unique number and this number
should be noted in every data collection form.
C.1 Hearing examination for children (age up to 5 years): otoacoustic emission (OAE)
testing should be undertaken in children up to 5 years of age.
C.1.1 Type of OAE test to be used: While either Distortion Product Otoacoustic Emissions
(DPOAE) or Transient Evoked Otoacoustic Emissions TEOAE can be used for hearing
screening, it is recommended to use DPOAE as part of this survey protocol, for
consistency.
C.1.2 Test procedure: The test procedure will depend on the equipment being used. It is
important for the testers to familiarize themselves with the relevant equipment and to
follow the relevant test procedure.
C.1.3 Pass/refer criteria: will depend on the equipment being used. It is important for the
testers to familiarize themselves with the relevant equipment and its pass/refer
criteria
C.1.4 Recording results: the test result should be recorded as a “pass” or “refer” for each
ear separately. If the child has a “pass” result in both ears, this should be recorded.
The final of the two results should be recorded in the data collection form (separately
for each ear) as 1=pass; 2=refer; 3=could not be performed. If for any reason the test
cannot be administered, or if the child fails the OAE test in either ear, the child should
6 Validated sound level apps can be used such as: NIOSH SLM, SPLnFFT, Noise Hunter, NoiSee or SoundMeter
(http://scitation.aip.org/content/asa/journal/jasa/135/4/10.1121/1.4865269).
48
be re-evaluated on another day. When this is not practical, the repeat OAE test should
be attempted the same day.
Note: When the child repeatedly fails the OAE test, the child must be referred for an
Auditory Brainstem Response test (ABR) which should be done for diagnostics
purposes.
C.2 Audiometry (age 5 years and over)
Pure tone audiometry is performed for all individuals who are 5 years of age and over.
The system and method must be uniform throughout each survey. The hearing
examination procedure should be carefully explained to each person. Although every
effort should be made to use pure tone audiometry in children aged 5 years and over,
if a child cannot be tested this way, OAE test (section BI) can be used.
C.2.1 Ambient noise See previous section.
Calibration The audiometer needs to be calibrated regularly to ensure
reliability. Portable audiometers should be calibrated as follows:
1. Daily biologic calibration check.
2. Regular audiometric calibration, at least every 6
months.
3. Regular battery testing to make sure they are at full
strength.
4. If an audiometer requiring power connection is
used, a voltage stabilizer should be used since reduction
in
voltage can produce reduction in audiometer output
Testing procedure The person who being tested should be seated with his/her
back to the control panel and examiner. Each ear should be
tested separately, the right ear first. If possible, noise-excluding
headsets should be used. The headphones should be well
fitting. Each time the participant hears the sound, he/she
should respond to the tester by using a simple sign, e.g. raising
a hand.
C.2.2 Hearing
thresholds
0.5, 1, 2, 4, 6 and 8
KHz
Therefore, the presentation of sound at the start of the test
should be 60 dBHL at 1 KHz. If there is no response to this
threshold, it should be increased in 10 dB steps until the
participant responds to the sound. Once the participant has
heard a sound, the threshold of hearing should then be
established by decreasing thresholds by 10 dB steps and
49
increasing by 5 dB steps until the threshold is established by
participant confirming threshold on three successive occasions.
No correction factors should be used since these will be
inserted at the analysis stage if necessary.
These thresholds should be established in the same manner at
0.5, 2, 4, 6 and 8 KHz and then finally, the threshold should be
established again at 1 KHz and should be within 5 dB of the
original measurement at 1 KHz. If not, repeat the whole
audiogram.
Hearing thresholds should be entered for each frequency in
both ears separately in the data collection form. Any threshold
above 20 dB HL is considered indicative of hearing loss.
C.2.3 Screening
examiner number
Each screening examiner should enter his/her assigned code.
C.2.4 NOT FULLY
EXAMINED IN
THIS SECTION
This box should be ticked if the participant was not fully
examined in Section C.
C.2.5 SCREENING
EXAMINER’S
REMARKS
This is the opinion of the examiner as to why the Examination
could not be completed satisfactorily. The reason should be
recorded. The equipment number should be noted here.
50
D: EAR EXAMINATION
D.1 Auricle
1. Auricle malformation Examine the auricle to look for any malformations, such as
small sized or absent pinna, or presence of sinus/es in front
of the pinna. Enter the appropriate code:
1= malformation of auricle; 2=no malformation; 3=uncertain
D.2 External ear canal
1. Wax
[Removed]
1=yes; 2=no; 3=uncertain
Mark the appropriate box with a tick if the wax was removed
by any member of team*.
2. Foreign body
Removed]
1=yes; 2=no; 3=uncertain If identifiable, please describe
what it is in Special Examiner's Remarks.
Mark the appropriate box with a T if object is removed from
ear canal by member of team‡‡.
3. Inflammation
[Treated]
Is there any redness or tenderness of the ear canal
1=yes; 2=no; 3=uncertain
Mark the appropriate box with a “T” if treatment has been
given for the ear canal inflammation.
4. Otorrhea
Look for evidence of any pus in the ear canal. 1=yes; 2=no;
3=uncertain.
If the ear canal has been cleaned by a member of the team
so that the ear canal is visualized, mark the appropriate box
with a “T”. If the ear was not touched leave box blank.
5. Fungus Look for any evidence of a fungal infection. 1=yes; 2=no;
3=uncertain
Mark the appropriate box with a “T” if treatment has been
given for the fungus in ear canal.
‡‡ The special examiner for Section C may wish the subject to have repeat hearing thresholds for diagnostic purposes
after the removal of wax, or a foreign body, or pus. These repeat hearing thresholds should not be recorded in the
boxes in Section B.
51
. D.3 Tympanic membrane: Examine the tympanic membrane with the use of an otoscope and
look for evidence of any abnormality such as perforation, dullness or retraction, bulging
and red tympanic membrane. Enter the number corresponding to the findings, as given
below. If the tympanic membrane is normal or if it cannot be clearly examined, this should
be noted.
1. Perforation 1=yes; 2=no; 3=uncertain
2. Dullness or retraction If the tympanic membrane is dull or retracted and the light
reflex is poor, mark yes.
1=yes; 2=no; 3=uncertain
3. Bulging and red Look for evidence that the tympanic membrane is tense and
bulging and the drum mucosa appears red. These signs,
when seen together with ear pain are indicative of acute
otitis media.
1=yes; 2=no; 3=uncertain
4. Normal If the eardrum appears normal, mark yes
1=yes; 2=no; 3=uncertain
5. Not seen If the tympanic membrane cannot be examined respond yes:
1=yes; 2=no
D.4 Middle ear: In case of a perforation of tympanic membrane, examine the middle ear
mucosa through the perforation and note the findings as follows:
1. Otorrhea If, on otoscopy, there is definitely otorrhea within the
middle ear, enter 1=yes; 2=no; 3=uncertain
2. Cholesteatoma Examine for evidence of cholesteatoma
1=yes; 2=no; 3=uncertain
3. Normal If no abnormality is identified
1=yes; 2=no; 3=uncertain
4. Not seen If the middle ear cannot be seen for any reason, respond
yes: 1=yes; 2=no
52
D.5 Others If there is any other abnormal finding related to any part of the ear or mastoid
region and specify these findings in the space provided.
D.6 Tympanometry This should be conducted using a portable tympanometer. Participants
with canal inflammation or with impacted wax, blood, pus discharge, debris, foreign
body or growth in the external auditory canal should be excluded from this test. The
type of curve (A, B, C, AS or AD) should be determined and recorded as follows:
1=A; 2=B; 3=C; 4=As; 5=AD; 6=uncertain; 7=test could not be performed
D.7 Examiner number Each screening examiner should enter his/her assigned code.
D.8 Not fully examined in this section This box should be ticked if the participant was not
fully examined in Section D.
D.9 Examiner's remarks This is for the opinion of the examiner as to why the examination
could not be completed satisfactorily. The reasons should be recorded.
E: PROBABLE CAUSE OF HEARING LOSS
This section should only be filled for those participants where an abnormality was identified
during ear or hearing examination. If none is found, this section should be left blank.
The definition of normal hearing, for the purposes of this section, is that either in section C.2, the
OAE test result is pass in both ears; or that in section C.3 no hearing threshold box is marked as
26 dBHL or greater.
Common ear diseases that are encountered should be documented (and coded). The following
case definitions and codes should be applied. Examination of ear should be undertaken by an
experienced otoscopist and whenever possible, any diagnosis of ear conditions be made in
consultation with a special examiner trained in identification of ear problems. The survey
protocol does not aim to identify and document the cause of hearing loss in each person
identified as having hearing loss. It aims only to document those problems which are evident
upon examination of external ear and tympanic membrane.
E.1 None identified (unknown) 1=yes; 2=no
53
E.2 Impacted wax (wax) 1=yes; 2=no; 3=uncertain
In cases where wax is encountered in the ear canal, written consent for wax removal
should be obtained from the adult participant/child’s guardian before an attempt is
made by a trained person to remove it. If wax cannot be removed and it was
completely obstructing the view of the tympanic membrane, it should be noted as a
diagnosis of impacted wax. In those cases where the wax can be easily removed from
the canal with the help of a probe and the tympanic membrane is clearly visible, it is
not considered to be impacted and such cases should not be classified as impacted
wax.
E.3 Chronic suppurative otitis media (CSOM) 1=yes; 2=no; 3=uncertain
Recurrent or persistent ear discharge lasting for more than 2 weeks and the presence of
a central perforation are criteria for diagnosis of chronic suppurative otitis media of the
tubotympanic (safe) type.5 Active ear discharge is a requirement for this diagnosis.
History of ear discharge, with evidence of cholesteatoma, marginal or attic perforation
or evidence of retraction pocket, point to a diagnosis of atticoantral (unsafe) type of
CSOM.
E.4 Otitis media with effusion (OME) 1=yes; 2=no; 3=uncertain
Dull, retracted tympanic membrane with reduced mobility and/or evidence of fluid in
middle ear, along with a type B tympanogram are considered indicative of OME.
E.5 Acute otitis externa (AOE) 1=yes; 2=no; 3=uncertain
Signs of inflammation of the external auditory meatus with or without ear discharge and
fungal debris are indicative of otitis externa.
E.6 Acute suppurative otitis media (ASOM) 1=yes; 2=no; 3=uncertain
Diagnosis of ASOM is made on the basis of redness, bulging or acute discharging
perforation of the tympanic membrane, mostly with a history of acute ear pain.
E.7 Foreign body in the ear (FB) 1=yes; 2=no; 3=uncertain
This is the presence of any external object or creature in the ear.
54
E.8 Age-related hearing loss (ARHL) 1=yes; 2=no; 3=uncertain
It should be suspected as the probable cause in people above the age of 60 years with
no history of childhood hearing loss or other risk factors.
E.9 Any other: (to be noted) Examiner’s comments:
Comments from examiner regarding probable cause of hearing loss and any specific
factors worth noting.
F. Any further action required should be noted in consultation with a doctor trained to make the
decision for medication and referral.
55
SECTION 7: Protocol for rapid assessment of
hearing loss
(written by Tess Bright and Hannah Kuper,
The London School for Hygiene and Tropical Medicine)
A tool for rapid assessment of the prevalence of hearing loss is useful for settings where data
are needed quickly, or where time- or cost-related factors are barriers to carrying out a full
epidemiological survey. Rapid assessment provides a quicker and cheaper option that can
serve to:
provide an estimate of hearing loss prevalence in people aged 50 years or older;
give an indication regarding hearing loss prevalence in the community;
serve as a first step to the planning of, and raising resources for, an in-depth survey, as
outlined in this survey handbook.
This section provides an outline of the RAHL concept and methodology.
Rapid assessment: the concept
RAHL provides a cheaper and quicker alternative to a full epidemiological survey by:
• Restricting the sample to people aged 50 years and older: the age restriction is based on
evidence that over 75% of people with hearing loss are aged 50 or older. Due to the
higher prevalence of hearing loss in this group, the sample size needed is drastically
reduced.
• Simplified examination protocol: the use of validated smartphone/app-based tools
offers a cheaper and quicker means for prevalence assessment.
• Simplified two-stage sampling method.
• Simple standardized analyses.
56
AIM OF RAHL
To estimate the prevalence of hearing loss in people above 50 years and its probable causes
To make projections regarding the prevalence of hearing loss in the entire population
SURVEY DESIGN AND SAMPLE
Sample size is estimated locally based on the approximate prevalence of hearing loss in the
target population i.e. people above 50 years of age (see Section 1, subsection on “Estimating
sample size” for details of sample size estimation). Given the high prevalence of hearing loss in
this age group, a much smaller sample size is expected as compared to a full population-
based survey including all age groups.
Survey design: A cluster design is recommended (see Section 1, subsection on “Survey design
and methods”). Each cluster should comprise 50 sampling units (50 individuals above the age
of 50 years). The total number of clusters will depend on the sample size required based on
the expected prevalence.
Sampling strategy: A two-stage sampling procedure is recommended.
First, the required number of clusters will be selected through probability proportionate to size
(PPS) sampling using the most recent national census or similar source of population data. The
next step is to select households within clusters. This can be done by a compact segment
sampling method (see Box 11).
Each cluster should be divided into segments of approximately 50 people aged ≥ 50 years. For
example, if a cluster includes 200 people aged ≥ 50 years then it should be divided in to four
segments.
One segment should be chosen at random by drawing lots, and all households in the chosen
segment visited, until 50 people aged ≥ 50 years are identified. If the segment does not include
50 people aged ≥ 50 years, then another segment should be chosen at random and sampling
continued until 50 people are identified and examined.
Box 11: Compact segment sampling
57
EXAMINATION PROTOCOL
Data collection should be done by going from house to house and seeking out all household
members aged 50 years or above.
Following enumeration and obtaining basic demographic information for each participant, all
participants should undergo a hearing assessment. In the case of rapid assessment (in contrast
to the recommendations made in Section 3 of the survey protocol), all examinations should
take place in the participant’s home at the time of enumeration. There are two stages to a
hearing assessment under the Rapid Assessment Hearing Loss protocol: 1) hearing assessment;
2) assessment of probable causes of any hearing problems.
Step 1: Perform a hearing assessment
All eligible participants should be assessed by a trained data collector using validated,
automated audiometry software. The chosen system must be able to determine hearing
thresholds at 500–8000 Hz. Details of suitable software applications evaluated by WHO are
provided in Annex 3.
Background noise should be measured and recorded before each assessment either within the
mobile testing application (as some have inbuilt noise monitoring capabilities) or using a
separate sound-level meter app. Information is available on page no. 44 the data collection
form.
A pure tone average of thresholds at 0.5, 1, 2, 4, 6 and 8 kHz in the better ear should be
obtained to determine the hearing threshold and estimate any degree of hearing loss, in line
with the WHO definitions (see Section 4)
• Before the survey takes place, calibrate the chosen mobile-based audiometry.
Step 2: Assess probable causes of any hearing loss
All participants should be examined by a trained local clinician (audiologist, ENT
specialist or ENT clinical officer) using otoscopy. However, where this is not feasible,
video-otoscopy (can be smartphone-based) with remote interpretation is acceptable.
58
In addition, a questionnaire should be administered for those identified as having
hearing loss to assist in the determination of the probable causes (see Section B of the
data collection form).
SURVEY TEAM
A RAHL team should consist of the following members:
• Team leader
• Ear and hearing care workers (one of them may be team leader)
• Field workers
• Driver
It is also desirable to have one person with specialized knowledge about ear diseases and
hearing loss, such as an ENT specialist or audiologist, as part of the team. If possible, one
member of the team should be conversant in the use of local sign language
EQUIPMENT REQUIRED
Otoscope
Hearing testing software applications with required hardware
Other tools (such as ear probe, forceps etc.)
Disposables
Eardrops and other medicines
Tablet for data collection
Missed participants
Persons above the age of 50 years that could not be examined during the first visit must be
followed up and arrangements made to ensure their inclusion in the sample. This is especially
important as it may be more likely to find older individuals at home compared to other eligible
community members. This is essential in order to ensure that a representative sample is
included.
59
DATA ENTRY AND ANALYSIS Where possible, data collected should be entered directly into a mobile application (such as
Open Data Kit). Where not possible, data must be collected using paper forms. Data should be
double-entered, with the two versions compared to establish accuracy. Consistency checks
should then be run to ensure that data are feasible (e.g. no cause of hearing loss should be
included for people with normal hearing).
Data should be analyzed as recommended to allow comparison of data. The main study
outcomes for the RAHL are:
• ·Overall prevalence of hearing loss (mild, moderate, severe and profound) in persons
aged 50 years and above (with 95% confidence intervals). Prevalence of moderate or
higher grade of hearing loss in persons aged 50 years and above (with 95% confidence
intervals).
• Probable causes of hearing loss in study population.
• Data should be disaggregated by age (50–59; 60–69; 70–79; 80+; above 100), gender,
education, rural/urban locations and any other relevant parameter.
Data can also be extrapolated to provide prevalence estimates for the entire population using
the following steps (see Box 12 for an example):
Note the prevalence of hearing loss in population aged 50 years or above, determined
through the RAHL survey: a
Note the proportion of population aged 50 years or above, from available census or
population information: b
Note the population of interest (either 1 million people or for the defined geographic
population): Total population: c
Estimated proportion of hearing loss in those aged ≥ 50 (out of the total population)§§: 75%
Calculate the number of people aged ≥ 50 with hearing loss in the population (d) =a*b*c
§§ This can be estimated based on available literature regarding prevalence of hearing loss in the population. If such
information is not available, researchers can work with the assumption that the estimated proportion of hearing loss
attributable to the age group above 50 years is 75%. This is based on an analysis of available WHO data
(https://apps.who.int/iris/handle/10665/206141)
Box 12: Extrapolating estimated prevalence among persons aged 50+ to entire population: an example
Note the prevalence in population aged ≥ 50, determined through the RAHL survey: a = 0.15
Note the proportion of population aged ≥ 50, from available census or population information: b = 0.10
Note the population of interest (either for 1 million people or for the defined geographic population): c = 1 000 000
Estimated proportion of hearing loss in ≥ 50 (out of the total population): 75%.
Number of people aged ≥ 50 with hearing loss d = (0.15*0.10*1 000 000) = 15 000
Number of people with hearing loss in population e = 1/0.75 * (0.15*0.10*1 000 000) = 20 000
Prevalence of hearing loss in the population e/c = 20 000/1 000 000 = 2%
60
Calculate the total number of people with hearing loss in the population (e) = 1/0.75 * (a*b*c)
= 1/0.75*d
Estimate the prevalence of hearing loss in the population = e/c
ADVANTAGES OF RAHL
• Requires fewer resources and time.
• Provides a valid assessment of hearing loss prevalence in population aged ≥ 50.
• Provides an assessment of probable causes of hearing loss in population aged ≥ 50.
• Provides an estimate of hearing loss prevalence in the entire population.
LIMITATIONS OF RAHL
Children and adults aged below 50 years are not assessed in the survey, therefore enough
evidence needs to be gathered by comparing the RAHL tool and the survey protocol for all
age groups, carried out simultaneously in the same geopolitical areas, in order to be able to
extrapolate valid and reliable hearing loss prevalence estimations for each age group
(particularly for infants and children), based on the results obtained with the RAHL tool for the
50+ age population.
Unless care is taken, sampling may capture only those older adults that are at home, excluding
adults aged ≥ 50 years who are working.
61
ANNEX 1: Sample size estimation and sampling
This annex provides useful online resources to assist in sample size calculation, and complex
sampling techniques and weighting. It also provides WHO regional estimates of hearing loss
prevalence that can be used for sample size calculation, in absence of suitable local, national
or regional published estimates on hearing loss prevalence.
Available online resources
1. WHO STEPS: WHO’s STEPwise approach (“STEPS”) supports countries in computing
national/subnational prevalence estimates of non-communicable diseases (NCDs) and
their risks factors. STEPS comprises several practical resources and tools, including the
STEPS manual, which sets out the steps for accomplishing a national/subnational
survey. It includes sampling size guidelines, comprehensive explanations, and sampling
size determination examples (see pages 86 to 93). The resources include a sample size
calculation tool*** that can be easily applied to determine the sample size for a study to
determine hearing loss prevalence in the population.
The STEPS resources also include a sampling spreadsheet that can help biostatisticians
and epidemiologists to:
• determine the probability proportional to size (PPS) sampling
• undertake simple random sampling
• weight the data.
Steps of the sampling procedure and related examples are provided on pages 85 to 113
of the manual.
*** STEPS sample size calculator and sampling spreadsheet is available at
http://www.who.int/ncds/surveillance/steps/resources/sampling/en/, accessed 3 October 2019.
62
2. Demographic and Health Surveys Program (see DHS) collects and disseminates
nationally/regional representative data on several health and related topics, and has
developed several tools and resources to undertake a national survey (see survey data).
It includes a comprehensive Survey Household Manual which explains the steps for
performing a national/subnational survey, including sample size calculation, and
examples (see pages 7 to 12); sampling techniques with excel templates for stratified
PPS sampling or equal probability systematic sampling (pages 52 to 70); and a full
glossary of complex sampling terms (see pages 80 to 87).
3. The Demographic and Social Statistics Division of the UN (see UNSD) collects, compiles
and disseminates official statistics on a wide range of social and demographic topics,
and develops standards, guidelines and methods. UNSD has developed several tools
and resources to accomplish effectively a national survey (see survey’s publications),
including a comprehensive Survey Household Manual. The manual explains the steps
for performing a national/subnational household survey, including sample size
calculation, and examples (pages 34 to 42); as well as sampling techniques and
examples for stratified PPS sampling or equal probability systematic sampling (see
pages 25 to 72).
WHO regional estimates
In 2012, WHO developed estimates of the number of people with hearing loss globally.
These estimates are based on a meta-analysis of population-based studies of hearing
loss prevalence in different countries, and were adjusted for population changes in
2018 (available at https://www.who.int/deafness/estimates/en/).
Table 1.1 provides the key prevalence statistics in different regions, while Table 2 provides a list
of the countries included in each region. Sample size estimation requires an initial estimate of
hearing loss prevalence. Ideally such a figure should be sought in published peer reviewed
literature that provides a valid assessment of hearing loss prevalence in the target population.
Alternately, such data can be obtained through a pilot study in the target population. Where
none of these is available, data from sources listed below can be used, in consultation with a
biostatistician for estimation of sample size, directly using the formula provided in Section 1,
63
“Survey design and methods” subsection of the handbook, or through the sample size
calculation tool.
Sources of global and regional hearing loss estimates
1. WHO global and regional estimates on prevalence of hearing loss (these are
summarized in Table 1.1 and Table 1.2) Fuller estimates are available at
https://www.who.int/deafness/estimates/en/).
2. The Institute for Health Metrics and Evaluation (IHME) GBD
Global/Regional/Country estimates on prevalence of hearing loss (choose context
impairment and impairment hearing loss under the “estimates GHDx tool”).
3. Stevens G, Flaxman S, Brunskill E, Mascarenhas M, Mathers CD. Global and regional
hearing impairment prevalence: an analysis of 42 studies in 29 countries. European
Journal of Public Health. 2013;23(1):146–15.
64
Table 1.1: WHO 2018 estimates for prevalence of disabling hearing loss (DHL)* in regions of the world
Prevalence
of DHL in
the world
%)
Prevalence
of DHL in
high-
income
countries
(%)
Prevalence of
DHL in the
Central/Eastern
Europe,
Central Asia
region (%)
Prevalence
of DHL in
the Sub-
Saharan
Africa
region (%)
Prevalence
of DHL in
the
Middle-
East and
North
Africa
region (%)
Prevalence
of DHL in
the South
Asia
region (%)
Prevalence
of DHL in
the Asia-
Pacific
region (%)
Prevalence
of DHL in
the Latin
America
and
Caribbean
region (%)
Prevalence
of DHL in
the East
Asia
region (%)
0–14 years total 1.73% 0.50% 1.55% 1.93% 0.92% 2.39% 2.02% 1.60% 1.28%
0–14 years male 2.13% 0.62% 1.91% 2.38% 1.13% 2.94% 2.49% 1.98% 1.56%
0–14 years female 1.31% 0.38% 1.16% 1.47% 0.69% 1.80% 1.53% 1.21% 0.95%
15–34 years total 1.73% 0.39% 1.30% 1.93% 0.79% 2.24% 1.84% 1.33% 1.15%
15–34 years male 2.13% 0.49% 1.60% 2.39% 0.98% 2.76% 2.27% 1.65% 1.40%
15–34 years female 1.31% 0.30% 0.98% 1.46% 0.59% 1.68% 1.40% 1.01% 0.86%
35–64 years total 6.53% 2.18% 6.55% 8.30% 3.69% 9.96% 8.42% 6.27% 5.84%
65
35–64 years male 7.92% 2.69% 7.96% 10.06% 4.48% 12.02% 10.22% 7.67% 7.10%
35–64 years female 5.13% 1.67% 5.28% 6.62% 2.83% 7.81% 6.68% 4.94% 4.53%
65 or more years total 33.17% 18.36% 36.56% 44.38% 26.40% 48.85% 43.69% 38.47% 33.71%
65 or more years male 37.73% 21.03% 41.47% 50.30% 30.82% 54.59% 49.24% 43.52% 38.28%
65 or more years female 29.43% 16.23% 33.73% 39.60% 22.58% 43.51% 39.46% 34.53% 29.50%
15 years or more (all adults) 7.64% 5.36% 9.95% 6.47% 4.08% 9.36% 8.62% 7.67% 8.03%
15 years or more (all adults male) 8.56% 5.82% 10.14% 7.41% 4.66% 10.69% 9.62% 8.50% 9.10%
15 years or more (all adults female) 6.72% 4.93% 9.79% 5.55% 3.48% 7.96% 7.65% 6.87% 6.93%
All ages total 6.12% 4.57% 8.36% 4.55% 3.17% 7.37% 6.90% 6.18% 6.85%
All ages male 6.86% 4.93% 8.45% 5.25% 3.64% 8.44% 7.71% 6.85% 7.72%
All ages females 5.36% 4.21% 8.27% 3.85% 2.67% 6.22% 6.10% 5.53% 5.93%
* Disabling hearing loss refers to hearing loss above 40 dB in the better hearing ear in adults and above 30 dB in the better hearing ear in children.
66
Table 1.2: Countries included in the regions listed in Table 1.1
East Asia region
Democratic People’s Republic of Korea, China (including Chinese Taipei, Hong Kong SAR (China),
Macau SAR (China)
Asia Pacific region
Cambodia, Cook Islands, Fiji, French Polynesia, Indonesia, Lao People’s Democratic Republic, Kiribati,
Malaysia, Maldives, Marshall Islands, Micronesia (Federated States of), Myanmar, Nauru, Niue. Palau,
Papua New Guinea, Philippines, Samoa, Solomon Islands, Sri Lanka, Thailand, Timor-Leste, Tonga,
Tuvalu, Vanuatu, Viet Nam
South Asia region
Afghanistan, Bangladesh, Bhutan, India, Nepal, Pakistan
Central / Eastern Europe and Central Asia region
Armenia, Azerbaijan, Georgia, Kazakhstan, Kyrgyzstan, Mongolia, Tajikistan, Turkmenistan, Uzbekistan,
Albania, Bosnia and Herzegovina, Bulgaria, Croatia, Czech Republic, Hungary, Macedonia (Former
Yugoslav Republic of), Montenegro, Poland, Romania, Serbia, Slovakia, Slovenia
Belarus, Estonia, Latvia, Lithuania, Moldova, Russian Federation, Ukraine
Middle East and North Africa region
Algeria, Bahrain, Egypt, Iran (Islamic Republic of), Iraq, Jordan, Kuwait, Lebanon, Libyan Arab
Jamahiriya, Morocco, Occupied Palestinian Territory, Oman, Qatar, Saudi Arabia, Syrian Arab Republic,
Tunisia, Turkey, United Arab Emirates, Yemen
Sub-Saharan Africa region
Angola, Benin, Botswana, Burkina Faso, Burundi, Cameroon, Cape Verde, Central African Republic,
Chad, Comoros, Congo, Côte d'Ivoire, Democratic Republic of the Congo, Djibouti, Equatorial Guinea,
Eritrea, Ethiopia, Gabon, Gambia, Ghana, Guinea, Guinea-Bissau, Kenya, Lesotho, Liberia,
Madagascar, Malawi, Mali, Mauritania, Mauritius, Mozambique, Namibia, Niger, Nigeria, Rwanda, São
Tomé and Príncipe, Senegal, Seychelles, Somalia, South Africa, Sierra Leone, Sudan, Swaziland, Togo,
Uganda, United Republic of Tanzania, Zambia, Zimbabwe
Latin America and Caribbean region
Argentina, Antigua and Barbuda, Bahamas, Barbados, Belize, Bermuda, Brazil, British Virgin Islands,
Bolivia, Chile, Colombia, Costa Rica, Cuba, Dominica, Dominican Republic, Ecuador, El Salvador,
Grenada, Guatemala, Guyana, Haiti, Honduras, Jamaica, Mexico, Netherlands Antilles, Nicaragua,
Panama, Paraguay, Peru, Puerto Rico, Saint Kitts and Nevis, Saint Lucia, Saint Vincent and the
Grenadines, Suriname, Trinidad and Tobago, Uruguay, Venezuela (Bolivarian Republic of)
High-income countries
Andorra, Australia, Austria, Belgium, Brunei Darussalam, Canada, Cyprus, Denmark, Finland, France,
Germany, Greece, Greenland, Iceland, Ireland, Israel, Italy, Japan, Republic of Korea, Luxembourg,
Malta, Monaco and San Marino, Netherlands, New Zealand, Norway, Portugal, Singapore, Spain,
Sweden, Switzerland, United Kingdom, United States of America
67
Complex sampling design data analysis software
The analysis of complex sampling design survey data has to be done using statistical software
that takes into account the characteristics of the sample design. These designs usually include
features such as multiple stages of sample selection, clustering, stratification, and unequal
probabilities of selection. The packages listed (alphabetically) here are those frequently used
to accomplish this task:
Epi Info is a public domain suite of interoperable software tools designed for the global
community of public health practitioners and researchers. You can use StatCalc calculator to
compute sample size (see samplesize). It can perform complex survey data manipulation (see
tutorial). More information about the software can be found at Epi Info.
R is an open-source software environment for statistical computing and graphics for several
research fields. You can use the samplesize4surveys library to compute sample size (see
samplesize4surveys). It can perform complex survey data manipulation (see tutorial). More
information about the software can be found at R Project.
SAS/STAT is a software suite developed by SAS Institute for advanced analytics, multivariate
analyses, business intelligence, data management, and predictive analytics. It can perform
complex survey data manipulation. Examples of such procedures are: surveymeans,
surveyfreq, surveylogistic, surveyreg, etc. (see tutorial). More information about the software
can be found at SAS/STAT.
Statistical Package for the Social Sciences (SPSS) software is a popular statistical package
developed by IBM that can perform complex data manipulation and analysis, such as sample
size calculator and complex sampling functions that take into account complex sampling
design. Examples of such procedures are: csplan, csselect, csdescriptives, csglm, etc. (see
tutorial). More information about the software can be found at SPSS.
Stata is a general-purpose statistical software package focused on research for data
management, statistical analysis, graphics, simulations, regression, and custom programming.
68
It can perform complex survey data manipulation. Examples of such procedures are:
svy:means, svy:regress, svy:logit etc. (see tutorial). More information about the software can be
found at STATA.
SUDAAN is a statistical software package for the analysis of correlated data, including
correlated data encountered in complex sample surveys. It can perform complex survey data
manipulation. Examples of such procedures are: crosstab, regress, logistic etc. (see tutorial).
More information about the software can be found at SUDAAN.
69
ANNEX 2: Minimum training requirements for non-
qualified audiologists to perform hearing
assessments
SCOPE
This annex outlines the minimum training required to perform an ear and hearing assessment
during the conduct of the prevalence survey using this handbook.
It is essential that staff assessing subjects during the survey should have clearly identified roles
and responsibilities and should receive appropriate training. This is vital for improving the
reliability of survey outcomes, and ensuring that subjects are informed and advised correctly of
their test results and intervention options.
Ideally, such training should be gained through accredited courses. However, where such trained
staff and accredited courses are not available, survey planners can use the guidance provided
here to train the survey staff. Training should be practical and include “hands-on” experience
under the supervision of qualified and experienced tutors.
Staff should also be trained to raise awareness as part of the survey’s activities. Training may
cover any cultural sensitivities specific to areas under coverage.
TOPICS TO BE COVERED IN TRAINING
1. Knowledge of ear and hearing loss (impairment)
• Structure and function of the ear, including how sounds travel along the auditory
pathway.
• The different types and levels/classification of hearing loss, and its common causes.
• Common ear problems and diseases, such as wax, otitis media, foreign bodies etc.
70
• Risk factors for hearing loss (impairment) and advice on preventative measures.
• Communication needs of hearing impaired people.
• Confidentiality, informed consent and record handling.
2. Testing procedures and limitations*
Pure-tone
audiometry
Principles of pure-
tone audiometry and a
recommended
procedure
Contraindications
for audiometry
Correct method for pure-
tone a-c threshold
determination. The
audiogram and
interpretation of results
Tympanometry Principles of
tympanometry
and a recommended
procedure
Contraindications
for tympanometry
Correct method for
tympanometry
The tympanogram and
interpretation of results
Otoscopy Theory and practice of
otoscopy
Relevant health
and safety issues,
e.g. discharging
ears
Interpretation of results,
effect of presence of wax on
results
*Trainees should be sensitized to factors that could affect the reliability or validity of the test
results.
3. Calibration of equipment, especially audiometers and tympanometers
4. Assessment of ambient noise levels using either:
• Sound level meters (validated apps for assessment of environmental sound levels e.g.
NIOSH NLM; SPLnFFT, Noise Hunter, NoiSee or SoundMeter).†††
5. Gaining informed consent and completing the data collection form (adults and children)
††† Available at http://scitation.aip.org/content/asa/journal/jasa/135/4/10.1121/1.4865269.
71
• Provision of required information regarding the survey, with the information sheet
provided.
• Gaining consent from subjects and parents/guardians.
• Recording participant’s history relating to ear and hearing problems using the data
collection form.
6. Data recording
• Accurate recording of data on the data collection form
• Input of data into database
7. Explanation of results and advice to patients
• Testers should be able to explain results of examinations and test procedures, and
advise on intervention options.
8. Referral procedures
Testers should be aware of when and where to refer patients (e.g. ENT, audiologist) with
consideration given to national and local protocols, and whether patients are already
receiving care for the condition found. A list of referable conditions and which practitioner
is most appropriate for a referral should be understood.
Resources
• BSA 2016 – Minimum Training Guidelines: Basic Audiometry & Tympanometry –
Minimum Training Guidelines. British Society of Audiology. www.thebsa.org.uk.
• Tolonen H (Ed.) European Health Examination Survey (EHES) Manual. Part A. Planning
and preparation of the survey. 2nd edition. Helsinki: National Institute for Health and
Welfare; 2016 (http://urn.fi/URN:ISBN:978-952-302-700-8, accessed 10 July 2019).
• Primary ear and hearing care training resource (Student’s workbook – intermediate
level). Geneva: World Health Organization; 2006
72
Annex 3: Tablet and Smartphone-Based Software
Applications for Hearing Assessment
Introduction
In 2015, WHO identified three smartphones- and tablet-based manual threshold pure-tone
audiometry (PTA) applications (apps) that could be used for hearing assessment in population
surveys. This was done initially through a review of available apps and portable equipment
verified in small studies.11 Two of these apps and their key features are listed in Table 1, with
permission from their developers.
11 Bright T, Pallawela D. Validated smartphone-based apps for ear and hearing assessments: a review. JMIR Rehabil Assist Technol.
2016;3(2):e13. Published 2016 Dec 23. doi:10.2196/rehab.6074.
Table 1: Summary of features
Frequencie
s tested
Decibe
l range
Pure
tone
s
Separat
e right
and left
side
testing
High
environment
al noise
warning
Software
compatibilit
y
Headsets
Audcal 0.5, 1, 2, 3,
4, 8 kHz
0–75
dB
Yes No Yes iOS To be used
with Apple
earbuds
hearTes
t
0.125–16
kHz
0–90
dB
Yes Yes Yes Android Comes with
calibrated
headphone
s
73
A study was undertaken to determine the validity and test-retest reliability of the three apps
by comparing these apps against conventional pure-tone audiometry using a diagnostic,
calibrated audiometer. The study was done in two different settings:
• Field setting: where hearing assessment was undertaken in a quiet room (not a sound
booth) using community health workers. Overall, 200 subjects were tested using the
apps.
• Clinical setting: where hearing assessment was undertaken in a sound booth and by
audiologists. A total of 50 subjects participated in this study.
Table 2 summarizes the findings of the study.
Table 2: Findings of WHO hearing app validation study
Test-retest reliability in field
and clinical settings
Validity12 in field setting
(testing in a quiet room)
Validity1 in clinical setting
(testing in a sound booth)
Audcal Very good Targets met Targets not met
hearTest Very good Targets met Targets met for all
frequencies except 6 kHz
Acknowledgement: We would like to express our gratitude to the software developers of these
apps for providing the required hardware and software for us to undertake our study.
12 Clinical targets for determining validity in field and clinical settings: 90% within 10 dB and 80% within 5 dB (Table 1) and across all
frequencies tested.
74
ANNEX 4: Exploring funding options – some tips
It is important to consider funding opportunities during survey planning. However, in many
countries it may be difficult to secure funding for a purely academic public health research. As
donors are increasingly interested in seeing real, measurable social impact and investing in
activities that lead to transformative and sustainable change, it is important to:
• be aware of the potential funder’s priorities, and to familiarize yourself with the types of
research projects they have supported in the past;
• research the potential donor and understand their funding interests and objectives. It is
essential to do initial background research first and if possible to speak directly with the
donor prior to applying; take time to really listen to the donor and their objectives, and
ask lots of questions.
• be flexible and open to adapting your proposal to the donor’s specific needs. At times
it may be relevant to add to the study aims and objectives depending on the priorities
of the potential funder. Additional objectives could include:
o access to services;
o investigating barriers to services;
o cost-effectiveness of screening/surveys/services.
However, there should be no flexibility in terms of the study methodology and its academic
rigour.
Framing the funding case
The ultimate objective of the prevalence survey is to generate actionable prevalence data that
can be used for advocacy and to develop evidence-based policies. In your funding proposal,
highlight the possible social impact and practical outcomes of the project.
75
Relevant information that should be included in the proposal include:
• The number of people that will be examined.
• The number of people likely to be referred to an ENT specialist/audiologist for further
testing.
• Services that will be provided through the survey, e.g. wax removal, treatment of acute
ear infections, hearing aid fitting among other things. Adding information on the
number of people likely to benefit from these services is important.
• Estimate of the project cost per person screened13 – if possible suggest ways to increase
the cost-effectiveness.
• Get your potential donor excited about the research project by highlighting the paucity
of prevalence data in the field of hearing loss. Explain that the prevalence study is a first
step for many countries to start increasing interventions, since without an accurate
estimate of hearing loss prevalence, governments struggle with policy and resources
allocation. As a researcher, you are proposing to provide that catalytic data to really
make an impact.
Potential donors:
• International aid agencies (starting with your own country)
• Research funding institutions
• Local government
• High net-worth individuals who are likely to have a personal interest in the field of
hearing loss
Outline of the funding proposal
Many donors will have their own format for making a funding proposal. However, where such
a format is not provided, the proposal can be structured using the following sections:
• Cover letter: a brief introduction to the project highlighting why it is necessary, and its
potential impact
• Title page
13 To obtain this number, simply divide your total budget by the number of people you plan to screen.
76
• Executive summary: one page summary of the need for, and goals of, the survey; its
potential impact and how it fits the funding agency’s priorities
• Institutional background: information about the institution leading the research,
including contact details of the institution and the principal investigator
• Introduction: a brief overview of the problem and how the survey would contribute to
addressing it
• Aims and objectives
• Implementation plan
• Implications for future policy development and research
• Budget
• Others: CVs, survey team, detailed protocol
The points listed under “Framing the funding case” should be kept in mind while drafting the
proposal.
Request in-kind contributions
Consider working with field partners that will share resources. For instance, local ENT doctors,
audiologists or other partners may consider providing their car free of charge. Some partners
may consider hosting international researchers in their homes, instead of having researchers
pay for extended stays in hotels. Local NGOs may wish to partner in the study and provide
local facilities and serve as part of the team without any financial compensation. The key to in-
kind contributions is to ask for help: in the new sharing economy, people are more open than
ever to share their resources for the common good.
Be resourceful
Ask your project team members how each person can individually help lower the overall costs
of conducting the survey.
Don’t give up
Fundraising can be a difficult process and rejection of a proposal is common. When a proposal
is rejected, always ask the donor about the reasons in as much detail as possible. This will help
improve your proposal and strategy. The most successful fundraisers are those that persevere.
77
ANNEX 5: Mobile application and web portal for
data collection
World Wide Hearing Foundation International (World Wide Hearing or WWH), a non-profit
organization that provides access to affordable hearing aids and hearing care, has created a
mobile data-collection application combined with a web portal for data aggregation to simplify
carrying out WHO hearing loss prevalence studies. The WWH app is Android-based, free of
charge and can be used both online and offline to capture patient data using the WHO
hearing loss data collection form.
By replacing the traditional paper-based data collection system with a mobile, Android-based
application, prevalence studies can be conducted more time- and cost-effectively by reducing
the need for manual data entry and transcription. Furthermore, the web-based administrator
portal enables researchers and/or project leads to track progress and data collection in real
time.
The WWHearing mobile application
The WWHearing mobile application uses the latest version of the WHO data collection form
for hearing loss prevalence studies. This data collection form includes the following sections:
• Consent to participate in the study
• Patient history questionnaire
• Hearing examination
• Tympanometry
• Ear examination
• Causes of hearing loss/recommended action questionnaires.
All of these sections can be filled out on a single Android device or by different team members
on multiple devices. If the devices have an active Internet connection during field tests, the
78
app will sync data between the different Android units, giving team members access to
previously recorded data from other devices.
If an Internet connection is not available during field tests, tested persons will be given the
same patient ID on different devices so that the datasets can be matched.
The hearing examination section gives three different options for data entry:
1. Otoacoustic emission (OAE) test for children up to 5 years of age.
2. For people aged 5 years and older:
o Manual data entry of hearing thresholds from any audiometer.
o For people aged 5 years and older:
o Automatic data entry using a third-party, Android-based hearing threshold
application. Currently, the WWH app can automatically integrate data collected
with hearTest.14
Administrator portal
After syncing the mobile devices online, data collection can be instantaneously tracked
through the web portal. For easy and prompt data analysis, collected raw data can be
downloaded daily from the data server.
The administrator web portal allows not only the tracking of the study’s progress, but also the
setting up of the study’s test clusters.
14 hearTest is a threshold-seeking hearing loss app. When using the hearing threshold app option, the
hearTest external app is launched and a hearing test can be performed. The results from this hearing test
app are automatically imported into the GHLD app. Any other test similar to hearTest could be used.
79
How to access the app and administrator web portal
Type Link/contact
WWH
app
This Android app can be downloaded through the Google Play Store at:
https://play.google.com/store/apps/details?id=com.qochealth.android.wwhi.ghld
Web
Portal
The Administrator web portal can be accessed at:
https://globalhearingloss.org
User
accounts
and info
For support, questions and creation of user accounts, please contact:
Global hearing loss database
The WWH app and administrative web portal are part of a larger initiative by World Wide
Hearing to create a Global Hearing Loss Database (GHLD). This is an open-access initiative that
aims to aggregate data across countries and regions for the public good, namely to help
increase hearing loss interventions, research, policy-making, and – more broadly – advance
the field of hearing loss.
80
ANNEX 6: Protocol for the validation of selected
software applications for assessment of hearing
thresholds
AIM OF THE PROTOCOL
This protocol aims to guide the validation process for any hearing testing software
applications identified for use during a hearing survey. This protocol was developed to test the
accuracy and reliability of three apps, but can be adapted for testing further apps, as required.
This protocol was developed by WHO, with contributions and support from Dr Peter Thorne
(University of Auckland, New Zealand); Dr Sharon Curhan (Harvard Medical School, United
States of America); and Dr Tess Bright (London School of Hygiene and Tropical Medicine,
United Kingdom).
STUDY OUTLINE
Overview: Assessment of accuracy15 and logistical challenges of identified hearing test software
applications (apps).16
Aim: To independently assess the validity and reliability of three different software apps on a
cohort of subjects using conventional pure-tone audiometry as the gold standard.17
Study questions
• How well does each app perform at determining hearing thresholds compared to gold
standard methods?
• How well does each perform with respect to reproducibility18 (test-retest reliability)?
15 Accuracy of a measurement: the degree of closeness between several measurements of a quantity and that quantity’s actual (true) value. 16 3 software applications have been identified in the literature (Shoebox audiometry, AudCAL, HearScreen). 17 Gold standard: Pure-tone audiometry conducted in a quiet space or a sound booth depending on the setting (see “Study setting” section). 18 Reproducibility of a measurement: the ability of a measurement to be replicated.
81
• Are the software applications valid tools for assessment of hearing thresholds in the
field19 compared with portable audiometry for the purpose of conducting ear and
hearing surveys?
Target
• Identification of hearing thresholds (in decibels) at 500 Hz, 1000 Hz, 2000 Hz, 4000 Hz,
(and if available, at 6000 Hz, 8000 Hz) and their average20 in each ear tested.
Study design
• Prospective diagnostic study
Study population
• Persons aged 4 years and older
Study setting
• Setting 1: Community. This assessment will be conducted at an identified place within a
village or block. This may be a community hall or school room engaged for the
purpose of this assessment.
o A portable audiometry unit21 will be set up, and carefully calibrated prior to
testing, to facilitate the conduct of the pure-tone audiometry.
o The ambient noise levels will be measured and testing will only be conducted if
the ambient noise levels are below a certain threshold (see next section on
“Ambient noise”).
• Setting 2: Clinic. This assessment will be conducted at an audiology clinic.
o Pure-tone audiometry and app testing will be conducted in a sound booth.
o The audiometer will be carefully calibrated prior to start of testing.
19 Field refers to a community setting as recommended in the Ear and Hearing Survey protocol. 20 Average hearing threshold will be calculated for frequencies of 500 Hz, 1000 Hz, 2000 Hz and 4000 Hz. 21 Portable audiometry unit: this equipment will be carefully calibrated and conducted in a quiet room.
82
Ambient noise
• Setting 1: Testing will be conducted in a quiet room and the ambient noise levels will be
below 40 dBA.22
o This level will be assessed using suitable sound-level measurement iOS-based
applications.23
o Each app has a mechanism of testing ambient noise and does not function in
noisy situations. This will be adhered to for each individual app.
• Setting 2: Testing will be conducted in an audiology clinic.
o The ambient noise levels will be measured and testing will only be conducted if
within American National Standards Institute (ANSI) standards.24
Sampling
• Consecutive: every person that meets the inclusion criteria will be included in the study
until the sample size is complete.
Inclusion criteria
• All persons aged 4 years and older that present for testing.
Exclusion criteria
• Congenital or acquired ear deformities that prevent adequate earbud insertion.
• Occluding cerumen or active ear drainage on otoscopy examination or any other
physical or mental disability where air conduction audiometry is not possible, and/or
the use of earphones is not possible.
• Visual impairment that precludes the use of index tests.
22 A 40 dBA level has been used based on the Occupational Safety and Health Administration (OSHA)
recommendations
(https://www.osha.gov/pls/oshaweb/owadisp.show_document?p_table=STANDARDS&p_id=9739) from the
United States Department of Labor and the American Academy of Audiology (AAA) Childhood Hearing
Screening Guidelines (http://www.audiology.org/publications-resources/document-library/pediatric-
diagnostics). 23 SPLnFFT, Noise Hunter, NoiSee or SoundMeter (http://scitation.aip.org/content/asa/journal/jasa/135/4/10.1121/1.4865269). 24 ANSI S3.1-1999 (R2003) American National Standard Maximum Permissible Ambient Noise Levels for Audiometric Test Rooms.
83
• Participants or legal guardian not willing to give consent for inclusion in the study.
• Participants aged under 4 years.
• Participants who are unable to follow simple commands, unable to remain alert during
the duration of the hearing test or unable to press the button on the touch-screen
device.
Sample size
• N = 12025
Who will conduct the testing?
1. All tests will be performed by an audiologist or audiology technician under the
supervision of an audiologist.
2. All 3 app-based hearing tests will be performed on all participants.
3. Some participants will undergo the app-based hearing tests twice (to determine test
reproducibility).
Test order
• There will be two designated testers26 (Tester 1 and Tester 2) performing the testing.
• There will be one group (designated as Group 1).
• Group 1: 120 participants will undergo testing with the 3 apps and pure-tone
audiometry.
o The index tests will be performed in random order by Tester 1.
o The pure-tone audiometry will be performed by Tester 2.
o The order of index tests and pure-tone audiometry will be randomized for each
patient.
o Tester 1 and Tester 2 will be blinded to each other’s results.
o The first 75 participants will undergo re-testing with one of the apps after at
least 10 minutes rest. This will be performed sequentially in order of subjects as
25 A number of 100 participants is the sample size needed to assess the accuracy within 7 dB of the measurement that gives 80% power at the 0.05 level of significance (two-paired sample t-test). We propose N = 120 for each app since we expect that in about 20% of the sample the data may not be collected according to the plan or at all (depleted phone batteries, environmental constraints not respected, etc.). 26 Testers will be audiologist or trained audiology technician under the supervision of an audiologist.
84
they are tested: app 1, app 2 and app 3, until each app has been re-tested 25
times each.
▪ For example, Participant 1 may test app 1, app 2, app 3, pure-tone
audiometry, and then app 1 again. Participant 2 may test app 1, app 3
pure-tone audiometry, app 2 and then app 2 again. Participant 3 may
test pure-tone audiometry, app 1, app 3, app 2 and then app 3 again.
This is to ascertain the reproducibility of testing for each app.27
o After 60 participants (i.e. half) have been tested, the testers will then switch roles
and Tester 1 will perform the pure-tone audiometry and Tester 2 will perform
the index tests.
Test procedure
• Each app will be tested in accordance with each app’s specific instructions.
• The appropriate headphone (or transducer) will be used in accordance with each app’s
specifications and calibrated as directed.
• Please refer to the relevant apps for further details on test procedure.
What will be recorded?
• The following will be recorded for all participants:
o Demographic data: name, date of birth, sex, contact information
▪ participants will be anonymized
o History of known ear disease:
▪ ear infections (yes/no)
▪ tympanic membrane perforation (yes/no)
▪ history of hearing loss (yes/no)
▪ history of tinnitus (yes/no)
▪ any other disease/s?
o Otoscopy examination performed by the audiologist prior to testing:
27 25 subjects will undergo testing with each app twice. This number is required for the “test-retest reliability”
assessment, to assess reproducibility of each app (another measure of precision).
85
▪ cerumen (yes/no). In cases where the cerumen is occluding the canal,
the subject will be excluded from the study
▪ tympanic membrane (TM) perforation with discharge (yes/no)
▪ TM perforation without discharge (yes/no)
▪ cholesteatoma (yes/no)
▪ canal atresia (yes/no)
▪ retracted TM
▪ other (describe):
o Ambient noise levels will be recorded and testing will only proceed if below a
specified level (40 dBA):
▪ if ambient noise levels are above this level, then testing shall not proceed
until it is below 40 dBA
o Hearing threshold with each of the app-based hearing tests, for each ear
separately. Hearing threshold will be determined separately in each ear, at the
following frequencies: 500 Hz, 1000 Hz, 2000 Hz and 4000 Hz (and 6000 Hz and
8000 Hz if possible).
o Type of earphone/headphone used for each test.
o Hearing threshold on pure-tone audiometry will be determined for each ear
separately at 500 Hz, 1000 Hz, 2000 Hz, 4000 Hz, 6000 Hz and 8000 Hz.
• The tester shall fill out a questionnaire on the proforma after each test for each
participant. This questionnaire will pertain to the ease of testing, time taken, issues
faced, battery life, interference due to noise etc.
Data
• For each participant, test results will be recorded for each ear separately.
• Data will be compiled and entered in the format provided.
• Data will be shared with WHO.
Data analysis
• Software for statistical analysis will be used.
• The following software (one or in combination) may be used in analysis: IBM SPSS
Version 20.0; SAS University Edition 2016 and MS Excel 2010.
• Other software for statistical analysis may be used.
86
• Mean and standard deviations should be calculated for each frequency. Calculation of
the mean differences between the threshold measurements obtained in each ear and at
each frequency should be calculated by pure-tone audiogram and each app.
o At least 90% of all threshold measurements obtained should be 10 dB or closer
than that obtained by pure-tone audiometry to be considered valid.
o At least 80% of all threshold measurements obtained should be 5 dB or closer
than that obtained by pure-tone audiometry to be considered valid.
o Further statistical analysis may be performed.
• Test-retest reproducibility would be assessed by comparison of thresholds obtained
from the first and second use of the app by the same subject in the same ear.
o At least 90% of all measurements obtained by each app should be within 5 dB
of each other.
o If the above results are not obtained, then the app would be deemed unsuitable
for use.
o Further statistical analysis may be performed.
Ethical issues
• The study must be approved by the Institutional Ethics Committee.
• Informed consent must be taken for all participants. For those below the age of 18
years, consent from a legal guardian will be required.
• All those identified as having hearing loss or ear disease during the study will be
referred for further assessment.
Test procedure for each participant
Step 1. Each participant will be registered and details on page 1 of the “Participant Proforma”
form filled out. If participants meet the inclusion criteria and are not excluded as per the
exclusion criteria, then an identification number will be assigned. This is to anonymize
participants in the study results.
Step 2. Each participant is then randomized to either be tested with the apps first or tested
with pure-tone audiometry first. After testing within either group, the participant then
87
proceeds to testing with the other test/s that have not been performed yet. Regarding testing
with the apps, the order of testing of the 3 apps shall be randomized from one participant to
the next.
Community setting
Step 2a. Testing with apps for setting 1 (community).
(i) Testing will be performed by an audiologist or audiology technician under the
supervision of an audiologist.
(ii) Testing shall take place in a quiet room. Ambient noise levels will be assessed using
suitable sound level measurement iOS-based applications28 and should be below 40
dBA. Testing will not proceed until levels are below 40 dBA.
(iii) Once the room is ready for testing, participants will be asked to sit down and their
identification number confirmed.
(iv) Headphones/transducer shall be placed on each participant and ensured that they are
appropriately fitted:
o AudCal: Apple headphones
o ShoeBox: TDH-50 (60 ohm), E-A-RTONE3A or BC-71 (Pro edition) (calibrated
headsets provided with the testing device will be used)
o HearScreen: Sennheiser HD 280 Pro (calibrated headsets provided with the
testing device will be used)
(v) Instructions given by each app will be followed in a step-wise manner to conduct each
test and data entered for each category in the participant proforma.
(vi) Testing will be repeated (up to 3 times) if the app registers that ambient levels are too
loud and the results are not valid. Steps to ensure a quiet room will be taken.
Step 2b. Testing with pure-tone audiometry for setting 1 (community).
(i) Testing will be performed by an audiologist or audiology technician under the
supervision of an audiologist.
28 Either SPLnFFT, Noise Hunter, NoiSee or SoundMeter (http://scitation.aip.org/content/asa/journal/jasa/135/4/10.1121/1.4865269).
88
(ii) Testing shall take place in a quiet room. Ambient noise levels will be assessed using
suitable sound level measurement iOS-based applications29 and should be below 40
dBA. Testing will not proceed until levels are below 40 dBA.
(iii) Once the room is ready for testing, participants will be asked to sit down and their
identification number confirmed.
(iv) Calibrated headphones shall be placed on each participant with a check to ensure
they are appropriately fitted.
(v) Pure-tone audiometry will be conducted in accordance with standard protocol for
identification of hearing thresholds as set out by the International Standards
Organization (ISO).30
Clinical setting
Step 2a. Testing with apps for setting 1 (clinic).
(i) Testing will be performed by an audiologist or audiology technician under the
supervision of an audiologist.
(ii) Testing shall take place in a sound booth with noise levels within ANSI
recommendations.
(iii) Participants will be asked to sit down and their identification number confirmed.
(iv) Headphones/transducer shall be placed on each participant with a check to ensure
they are appropriately fitted.
a. AudCal: Apple headphones
b. ShoeBox: TDH-50 (60 ohm), E-A-RTONE3A or BC-71 (Pro edition) (calibrated
headsets provided with the testing device will be used)
c. HearScreen: Sennheiser HD 280 Pro (calibrated headsets provided with the
testing device will be used)
(v) Instructions given by each app will be followed in a step-wise manner to conduct
each test and data entered for each category.
29 Either SPLnFFT, Noise Hunter, NoiSee or SoundMeter (http://scitation.aip.org/content/asa/journal/jasa/135/4/10.1121/1.4865269).
30 ISO 8253-1:2010.
89
(vi) Testing will be repeated (up to 3 times) if the app registers that ambient levels are
too loud and the results are not valid. Steps to ensure a quiet room will be taken.
Step 2b. Testing with pure-tone audiometry for setting 2 (clinic).
(i) Testing will be performed by an audiologist or audiology technician under the
supervision of an audiologist.
(ii) Testing shall take place in a sound booth with noise levels within ANSI
recommendations.
(iii) Participants will be asked to sit down and their identification number confirmed.
(iv) Calibrated headphones shall be placed on each participant with a check to ensure
they are appropriately fitted.
(v) Pure-tone audiometry will be conducted in accordance with standard protocol for
identification of hearing thresholds as set out by the International Standards
Organization (ISO).31
Step 3. After testing of the apps and pure-tone audiometry is performed and data collected,
the first 75 participants within the study shall undergo repeat testing with one of the apps
only. Each app will be re-tested 25 times and data recorded for each participant.
Step 4. After each test is performed, a short questionnaire shall be completed on the proforma
for each participant. After test results are collected for each participant, data shall be entered
into an Excel spreadsheet provided by WHO and this will be shared with WHO
31 ISO 8253-1:2010
Q.1. Time taken to perform test (in minutes and seconds): Q.2. Were there any special requirements?
Q.3. Any hindering factors? (e.g. short battery time, noisy environment, etc) Q.4. Any complaints or concerns from the subject?
Q.5. Any other comments?
Participant Proforma for Evaluation of Study Apps
Participant details
Name:
Age:
Gender: Male/Female (please circle)
Address:
Contact number:
Date/s of testing
History History of hearing loss: Yes/No History of ear infections: Yes/No History of tinnitus Yes/No History of tympanic membrane (TM) perforations:
Yes/No
History of ear disease: (please specify)
Duration: Symptoms:
Ear Examination
Cerumen: Yes/No , If occluding, then patient is excluded from study Perforation with discharge: Yes/No Dry perforation: Yes/No Cholesteatoma: Yes/No Canal atresia: Yes/No Retracted TM: Yes/No Any other (describe):
Frequency (Hz) 500 1000 2000 4000 6000 (if applicable) 8000 (if applicable)
Hearing Threshold Right Ear(dB)
Hearing Threshold Left Ear (dB)
If no exclusion criteria ticked above, then please assign participant IDENTIFICATION NUMBER (ID No.) here and proceed to testing: ID no. …………………………… Participant Identification Number:…………………………. Results for App ………………………………………………………………… (please fill in the app name) Was the ambient noise level (as assessed by the app*) too loud at one or more frequencies? Yes/No Each app detects the ambient noise level and can indicate if it is too noisy for suitable conduct of the test. The ambient noise levels should be below the indicated levels during the test. In case the test is conducted in a noisy environment, this should be indicated in the results.
* Each app detects the ambient noise level and can indicate if it is too noisy for suitable conduct of the test. The ambient noise levels should be below the indicated levels during the test. In case the test is conducted in a noisy environment, this should be indicated in the results
Exclusion Criteria: (please tick box if present)
• Congenital or acquired ear deformities that prevent adequate earbud insertion ☐
Occluding cerumen or active ear drainage on otoscopy examination ☐
• Any other physical or mental disability where air conduction audiometry is not possible, and/or the use of
earphones is not possible. ☐
• Visual impairment which is inconsistent with the use of index tests ☐
• Participants or legal guardian not willing to give consent for inclusion in the study ☐
• Participants aged less than four years ☐
Participants who are unable to follow simple commands, unable to remain alert during the duration of the hearing
test or unable to press the button on the touch-screen device
Please fill the following questions out before participant proceeds to next step.
Results for App …………………………………………………………………. (please fill in the app name) Was the ambient noise level (as assessed by the app) too loud at one or more frequencies? Yes/No
Please fill the following questions out before participant proceeds to next step.
Q.1. Time taken to perform test (in minutes and seconds): Q.2. Were there any special requirements?
Q.3. Any hindering factors? (e.g. short battery time, noisy environment, etc) Q.4. Any complaints or concerns from the subject?
Q.5. Any other comments?
Q.1. Time taken to perform test (in minutes and seconds): Q.2. Were there any special requirements?
Q.3. Any hindering factors? (e.g. short battery time, noisy environment, etc) Q.4. Any complaints or concerns from the
subject?
Q.5. Any other comments?
Frequency (Hz) 500 1000 2000 4000 6000 (if applicable) 8000 (if applicable)
Hearing Threshold Right Ear(dB)
Hearing Threshold Left Ear (dB)
Frequency (Hz) 500 1000 2000 4000 6000 (if applicable) 8000 (if applicable)
Hearing Threshold Right Ear(dB)
Hearing Threshold Left Ear (dB)
Pure Tone Audiometry Results Participant Identification Number:………………………….
Frequency (Hz) 500 1000 2000 4000 6000 8000
Hearing Threshold Right Ear(dB)
Hearing Threshold Left Ear (dB)
Note: this section is applicable to first 75 participants only
Repeat test results for App …..(please fill in the app name) Participant Identification Number:………………………….
Frequency (Hz) 500 1000 2000 4000 6000 8000
Hearing Threshold Right Ear(dB)
Hearing Threshold Left Ear (dB)
Q.1. Time taken to perform test (in minutes and seconds): Q.2. Were there any special requirements?
Q.3. Any hindering factors? (e.g. short battery time, noisy environment, etc) Q.4. Any complaints or concerns from the subject?
Q.5. Any other comments?
Q.1. Time taken to perform test (in minutes and seconds): Q.2. Were there any special requirements?
Q.3. Any hindering factors? (e.g. short battery time, noisy environment, etc) Q.4. Any complaints or concerns from the
subject?
Q.5. Any other comments?
World Health Organization
Department of Noncommunicable Diseases
20 Avenue Appia
1211 Geneva 27
Switzerland
https://www.who.int/health-topics/hearing-loss